Many critics of Bjrn Lomborg's book refer to Mark Twain's comment about "lies, damn lies and statistics," but I am more reminded of H. L. Mencken's remark, "For every problem, there is a neat, simple solution, and it is always wrong." The story of The Skeptical Environmentalist is one of a political scientist who wades into the vastly complex, unsettled literature of environmental science, scrutinizes a fraction of what is to be found there, and emerges confident that the simple summary he has developed is a fair and accurate representation of the sciencenotwithstanding the warnings of experts in the disciplines he skims that he is mistaken.

Lomborg has not taken these criticisms lying down, however, as his detailed rebuttal to our feature in Scientific American attests. Because of demands on their own time, most of the authors of our "Misleading Math about the Earth" section were unfortunately not available to write full responses to Lomborg's reply themselves. However, I have taken it upon myself to answer many of his counter-arguments, which display the same patterns of misdirection, bias and selective citation that our authors found in his book. (The responsibility for what appears here is entirely mine, but I do appreciate the contributions that our authors were able to offer in some sections, as noted.)


Lomborg's criticisms begin with the subtitle for our article, "Science defends itself against The Skeptical Environmentalist." As he insists, his book "makes a claim to science and to be factually based," and therefore it is inappropriate to describe his work as an attack on science.

Yet Lomborg does emphatically deride the conclusions, judgment and integrity of researchers and organizations whose work represents the mainstream of environmental science (a few examples are described below, and the book contains many others). As for the status of Lomborg's book: Many people claim to have scientific arguments when they do not: creationists, ancient astronaut theorists and the like. Being able to cite the scientific literature does not automatically confer validity on one's argument, nor does it make the argument genuinely scientific.

This is why Lomborg is wrong in his rebuttal when he writes, "The discussion is whether the statements in my book are correct or not." The discussion is not about whether his statements are correct; it is about whether his arguments are correctthe plans of thought he develops from those statements. Selective citations of the literature allow Lomborg to make statements that are correct but unrepresentative of the best, full state of environmental scientific knowledge. If Lomborg's book is science, then it should be judged by the criteria for other scientific publications. As many scientists have commented, his discussions would be unlikely to pass peer review because of his apparent unfamiliarity with prior literature, his selective citation of relevant work, and other faults.

The discussion is not about whether his statements are correct; it is about whether his arguments are correctthe plans of thought he develops from those statements.

Lomborg takes exception to my chastising him for "literally not seeing the forests for the trees." He writes that "the longest data series actually tells us of very little change in the world forested area," but this is misleading. My comment springs from the fact that the data series may describe little change in the forested area, but the actual forests are subject to considerable clearing and replanting. Lomborg thus treats forests of new trees as ecologically equivalent to old-growth forests, which is clearly not true.

But even on its own terms, Lomborg's statement about forest cover does not hold up well. He is relying principally on data from the U.N.'s Food and Agriculture Organization (FAO). As Emily Matthews of the World Resources Institute has pointed out, the FAO compiled that data for agricultural purposes, and eventually stopped doing so because it felt that the data were inadequate for assessing forests accurately. Lomborg also quotes FAO figures for tropical deforestation as a percentage of total global forest cover, which greatly understating the pressure on tropical forests, where most of the deforestation is taking place.

In his rebuttal, Lomborg next begs to be understood as trying to promote a properly balanced dialogue on environmental issues; when it comes to scientists, "we should take the science of these people seriously, but we should not uncritically adopt their evaluation of the problems." As he also articulates in his book, environmental policies should be settled after considering the perspectives and information from all segments of society: not just scientists but also economists, business leaders, political leaders, and all the rest.

This is such an eminently reasonable, desirable goal that a reader should be forgiven for forgetting that this dialogue already takes place. Perhaps circumstances are different in Lomborg's home of Denmark, but certainly in the U.S., no important environmental policies are set without considerable debate involving exactly the economic interests that Lomborg describes. Look at the ongoing controversies over drilling for oil in the Alaskan National Wildlife Refuge, over post-Kyoto greenhouse gas policies, over fuel efficiency standards for automobiles, over asbestos levels in drinking water. On every one of these issues, environmental scientists' evaluations of the problems have been interrogatedas they should be. Various interests may not be pleased with how these debates are settled, but it is a fantasy to pretend that environmentalists are simply dictating policies, or that the voices of economists are not being heard and often heeded.

The Skeptical Environmentalist thus presents itself as a disinterested inquiry into environmental science, but its implicit goal is to diminish the undue influence that Lomborg thinks environmentalists and scientists have, under the cover of restoring balance to the debate. As such, it is not really a scientific work so much as a polemic.

Lomborg is correct that the experts we invited were not chosen randomly. We chose Stephen Schneider, John Holdren, John Bongaarts and Thomas Lovejoy because they are leading spokespeople for the science that Lomborg criticizes. Why would we not give them a chance to reply, particularly when they are strong representatives of their fields?

Lomborg repeatedly brings up past statements of some these authors, presumably to ridicule them for their mistakes. Of course, the statements he unearths are frequently more than 20 years old. Scientists do sometimes make mistakes; they also revise their conclusions as new evidence accumulates. Unlike Lomborg, I see no reason for this to permanently discredit them.

He particularly seems to enjoy quoting a statement that Schneider made in an interview in 1989 about the "double ethical bind" researchers can find themselves in, to the effect that sometimes they might need to "offer up scary scenarios" to build public support. Lomborg does not overtly accuse Schneider of lying about global warming forecasts but his innuendo is clear.

Lomborg's use of this quotation is itself an interesting case of taking a statement out of context. Ironically, as Schneider has explained in the past, those words are drawn from comments he was making about the severe difficulty of explaining complicated scientific issued to the public through popular media that favor sound bite presentations. Communicating effectively and honestly almost inevitably means simplifying controversial scientific issues and using metaphors. Schneider explains this same point in his book Global Warming: "There is no simple formula for resolving the dilemma of balancing effectiveness against full disclosure, for one scientist's clear simplification could well be another's irresponsible oversimplification. Each tries to find the best path across this treacherous ethical ground." Lomborg does his best to make Schneider's quotation an admission of dishonesty, but I note that Lomborg also usually leaves off (or buries in a footnote) the last part of Schneider's quoted comment: "Each of us has to decide what the right balance is between being effective and being honest. I hope that means being both."

Springing from his selective quotation of Schneider, Lomborg goes on to question the integrity of Scientific American in commissioning this article: "The obvious lack of any concern for presenting a balanced review of my work calls into question the real purpose of this Scientific American feature....Could John Rennie have taken [Schneider's quote] as editorial advice?" (An aside: some of Lomborg's defenders have characterized our authors' remarks as being no more than "ad hominem assaults." But Lomborg attacks the character of our authors, this magazine and me far more personally than we do him.)

Given that Lomborg's work has few defenders within the sphere of environmental scientists, publishing what he would have considered a "balanced review" would have misrepresented how poorly Lomborg's book is regarded by scientists. More to the point, my introduction labeled our feature as a response by the scientific community that feels Lomborg misrepresents their work. It was not a disinterested review; it was not a debate. Magazines, even science magazines, publish articles with points of view. We then give people with opposing points of view an opportunity to reply, typically with a letter to the editors. We have followed that practice with Lomborg, who has certainly found other magazines such as The Economist that are more than willing to champion his side.

If Lomborg thinks there is a "glaring contradiction" in my writing that his book might hold "some truth" while also being "a failure" at proving its argument, then I am not sure how to make his thinking less black-and-white.


Lomborg seems to think that Schneider's first few paragraphs amount to an ad hominem dismissal of him. We disagree. Lomborg's lack of credentials as an environmental scientist does not disqualify him from making an environmental argument, but it does legitimately bear on whether his knowledge of the field and his paraphrasing of its findings can be trusted. Specifically, it raises the question of whether he relies on a small, biased base of data in drawing conclusions (which is precisely the criticism raised against The Skeptical Environmentalist). Fundamentally, Lomborg takes exception to Schneider's critical rhetoric, but of course Lomborg's own writing is no less skillful or fervent in attacking Schneider's integrity.

Schneider makes the correct point that sheer bulk of references does not prove that Lomborg's chapter on global warming has much validity. (Readers familiar with Vladimir Nabokov's novel Pale Fire know this well; at a more humble level, one could check Steve Mirsky's Anti Gravity column in the April 2002 issue of Scientific American.) While tut-tutting his critics' lack of footnotes, Lomborg trumpets his abundant use of them "since it really makes it a lot easier for my critics to attempt to show exactly where I might be wrong." Yet his critics are rarely attacking the facts in his citations per se; they fault what Lomborg does with them and, of course, his neglect of more relevant citations elsewhere.

Lomborg's lack of credentials as an environmental scientist does not disqualify him from making an environmental argument, but it does legitimately bear on whether his knowledge of the field and his paraphrasing of its findings can be trusted.

Next, let's turn to the subject of probabilities. Schneider faults Lomborg for failing to quantify what he means when he says "plausible." Lomborg's reply ignores this criticism and instead redirects attention to the Intergovernmental Panel on Climate Change's use of probability estimatessaying, in effect, that they are not really objective after all.

Of course, all predictions have an inherently subjective aspect to them because they are based on some assumptions. (We can predict with high certainty that the sun will rise to morrow because we assume that our knowledge of celestial mechanics is not wrong and that no unexpected phenomenon will obliterate the earth or sun over the next 24 hours.) The confidence levels embodied by the IPCC scientists' probability estimates are ba sed in most cases on empirical data: how well various models hold up during tests, how frequently different climate scenarios converge on the same outcomes, and so on. For Lomborg to spurn the IPCC estimates of probabilities, therefore, disrespects the efforts that scientists have put into quantifying possible outcomes of global warming. Moreover, Lomborg's argument still does not justify his failure to quantify his use of "plausible," which may mean no more than his opinion.

Against the charge that he generally dismisses IPCC ranges and focuses on the least serious outcomes, Lomborg says that he does write out the ranges from the main scenarios for sea level increase and for agricultural impact. True, but Schneider maintains that these examples mean less than meets the eye. The effect of global warming on agriculture, according to IPCC estimates, is on balance likely to be somewhat beneficialgood news on the face of it, but the particulars associated with that generalization need to be considered, too. The IPCC report noted that this gain is likely to occur only within the lower range of global warming estimates, and that even then, some net losses in agricultural productivity were probable for tropical countries.

Lomborg's recitation of the IPCC projections for the range of sea level increases mentions that they are relatively small. Yet the IPCC also notes that this small increase applies only to the 21st century. Long after carbon dioxide levels plateau, sea levels may continue to rise as the deep oceans continue to get warmer for centuries to come.

According to Schneider, Lomborg essentially ignores the damage that could be caused by severe droughts and storms within the range of IPCC scenarios. Lomborg also ignores the damage that would be done to nature itself, since he attaches no particular value to degradation of the environment except in so far as its products are traded in markets.

Schneider wrote that Lomborg does not include any probability estimates for serious outcomes. In this, he is mistaken, as Lomborg says: The Skeptical Environmentalist does include probability estimates associated with the West Antarctic Ice Sheet melting and with the thermohaline circulation shutdown. This overstatement aside, Schneider and other scientists still maintain that Lomborg pays inadequate attention to probabilities associated with most other serious outcomes.

Concerning Lomborg's description of a climate model described in a 1989 Hadley Center paper, Lomborg supports his paraphrasing of the work by citing a "major Science overview article." Readers familiar with Science as one of the most eminent journals in the world could be impressed by this referenceand misled. Unfortunately, it only highlights the slippery way in which Lomborg frequently mixes his citations from primary and secondary sources: the article in question is from Science, but it is a story written by a member of its news staff, not by a climate scientist. Science has outstanding news writers in its employ, but I do not think anyone at that publication would pretend that its news writers are held to the same standards of precision and accuracy that its scientist contributors are. Errors of this sort are also not likely to be caught and corrected with the same high scrupulousness. Nonscholarly articles of this type in Scienceor in Nature, or in Scientific Americanare simply not meant to be used as exhibits in a scholarly argument as Lomborg has tried to do.

Lomborg's rebuttal, in short, serves as a case study in how he picks and chooses unrepresentative citations to suit his ends.

Lomborg also asks why he should go back to the original research paper when more recent work has been done. He then cites a paper from 1993 to bolster his view (even though that paper also favors a higher range for likely climate change than his original citation did). What he does not mention, according to Schneider, is that the current versions of the model developed by the Hadley Center support the view that severe climate changes are strong possibilities. Moreover, if Lomborg favors using articles by Science's news staff as evidence, he ignores a recent article by the same news writer whom he quoted before ("Climate Change: Reducing Uncertainties of Global Warming," by Richard A. Kerr, in Science, vol. 295, pages 29-30, Jan. 4, 2002) that explains how recent studies indicate the upper ranges of climate sensitivity may be even higher than has been believed. Lomborg's rebuttal, in short, serves as a case study in how he picks and chooses unrepresentative citations to suit his ends.

Richard Lindzen, in a letter to Scientific American that Lomborg approvingly quotes, argues that Schneider explains Lindzen's work on the "iris effect" incorrectly; he goes on to say that "there remain uncertainties in our work, but Schneider's concern over 'extrapolation' is not one of them." In answer to that, it should first be acknowledged that Lindzen's doubts about global warming and the human role in it represent a tiny minority view, and that the evidence to which he points has not persuaded a large segment of the climatological community (see the profile of Lindzen in the November 2001 issue of Scientific American). Lindzen participated in both the IPCC and the National Academy of Sciences reviews of global climate change, but he failed to convince his colleagues that he was correct. In Schneider's view, Lindzen's work on the iris effect still relies on data collected over too small an area of the earth to be generalized reliably to a global level. The contribution of Lindzen's iris effect is only to widen the distribution of possible climate effects.

Furthermore, Schneider notes that a recent report from NASA/Langley ("The iris hypothesis: A negative or positive cloud feedback?" by B. Lin et al., Journal of Climate, vol. 15, pages 3-7, January 2002) suggests that even if the iris effect is real, the results on climate change may be worse than Lindzen predicts because of reflections from clouds; the calculated net effect would then cause slightly positive feedback (warming the climate) rather than strongly negative feedback (cooling it). Another new paper, from the University of Washington ("No evidence for iris," by D.L. Hartmann and M.L. Michelsen, Bulletin of the American Meteorological Society, Vol. 83, No. 2; February 2002), contests whether the iris effect occurs at all.

Schneider's point is not that Lindzen's work is right or wrong; such matters will continue to be debated and investigated by climate researchers. Rather, Lomborg seems to be giving undue attention to a controversial aspect of climate modeling because it favors his thesis.

Lomborg defends his mention of the Danish physicists' solar magnetic hypothesis because he means it as a "supplementary explanation," not an al ternative one. His justification only sidesteps Schneider's point, however. No one rejects the possibility that solar activity plays a role in temperature fluctuations. But it is not enough to point to fluctuations in solar activity, meteorological factors and temperature and then suggest a linkthat is the elementary mistake of seeing correlation as causation. As Schneider wrote, "The reason the IPCC discounts this theory is that its advocates have not demonstrated any radiative forcing sufficient to match that of much more parsimonious theories, such as anthropogenic forcing." In other words, no one can suggest a mechanism that adequately explains how the solar fluctuations could produce effects that would not be swamped by greenhouse gases and other human influences on climate. Lomborg may want to present the solar magnetic contribution as merely a supplement, but until demonstrated otherwise, it is a trivial supplement.

As Schneider noted in his article, Lomborg criticized the IPCC for dropping a paragraph from the Working Group II report that, in Schneider's words, "downgraded the significance of economic studies that aggregate climate change damages." In his reply, Lomborg repeats his opposition to the loss of this "clear message." Yet the text of his reply fails to grapple with Schneider's expressed point: that the message might have been clear, but it was misleading. As Schneider wrote, "The government representatives downgraded aggregate cost-benefit studies for a reason: these studies fail to consider so many categories of damages held to be important by political leaders as to render them just a guideline on market sector transactions."

Lomborg's reply does not offer any contradiction of this. Instead, he huffs that Schneider makes this point without reference (Schneider was part of the IPCC; he wrote in part from personal observation) and writes that "it seems highly unlikely that it would have been stopped for such a reasonit would have been a much more obvious choice to make the cost-benef it analysis more inclusive." More obvious, perhaps, but not necessarily feasible: some of those factors of importance to the government decision-makers are not easily quantifiable. Trying to fit them into an economic model, if yet possible at all, might only have yielded the sort of mushy economic analysis Lomborg decries elsewhere.

Rather than take Schneider's word for it, one can find this explanation in the IPCC report itself: "Application and extension of the economic paradigm certainly focuses attention of cost measures that are denominated in currency, but practitioners have been criticized on the grounds that these measures inadequately recognize non-market costs... Their list [of overlooked measures] includes monetary losses, loss of life, chan ges in the quality of life (including a need to migrate, conflict over resources, cultural diversity, loss of cultural heritage sites etc.), species or biodiversity loss and distributional equity" (Chapter 2, Working Group 2, section 2.5.6).

Lomborg writes, " is scientifically unreasonable to argue that since we don't have all the data, we shouldn't even try to get them but juststop!" But the report was supposed to summarize the best state of knowledge at the time of its completion. The deletion of the paragraph does not mean that science has ground to a stop on the issue. In the meantime, even if Lomborg resents that an economically favorable comment did not make it into the report, deleting the reference to cost-benefit analysis was wiser than including an unhelpfully incomplete one.

Against Schneider's criticism that Lomborg cites only one value for climate damages, Lomborg replies, "I merely write out the mean estimated cost of Nordhaus's latest model." But of course, this is Schneider's pointthat Lomborg chooses one number derived from one model without acknowledging the many other numbers from other models less favorable to his argument. Nordhaus's own publications produce a wide range of different values, for all that he chooses one as hi s aggregate result. Nordhaus's models have also been widely criticized for overstating the costs of climate-change mitigation while understating the benefits.

Lomborg further defends this figure by noting that it is roughly comparable to estimates from an earlier IPCC report, but in so doing he also overrules the IPCC's current lack of confidence in that number. Despite his protests, Lomborg is choosing one number out of a very large range and justifying it by citing selectively.

Concerning Schneider's point about "catastrophic outcomes," it appears that Lomborg and Schneider are arguing past one another because of the dual meanings of "catastrophe." Lomborg says, correctly, that most of the global warming analyses conducted by the IPCC and other s generally exclude global catastrophes such as the collapse of the West Antarctic Ice Sheet and the thermocline shutdown (as discussed above). In his article, however, Schneider is referring to the more common varieties of catastrophe that include killer floods and powerful hurricanes. Anyone who has lived though hurricanes such as Andrew and Mitch would probably accept "catastrophe" as a description of those events. And as Schneider writes, "it is precisely because the responsible scientific community cannot rule out such catastrophic outcomes at a high level of confidence that climate mitigation policies are seriously proposed."

To prove that he does not "ignore the findings of engineers," Lomborg points to two pages of discussion in his chapter. It would have been more accurate to say that Lomborg dismisses the findings of engineers. Economists do find fault with engineers' models, and vice versa. Rather than present this as an unsettled controversy, Lomborg takes the economists' side. He may well believe their models are better, but the short shrift that he gives to the controversy does a disservice to the uncertainties hanging over this subject. In any case, this is one more instance of Lomborg exhibiting a bias for the economists' views.

An other example appears in his writings about below-zero costs. Essentially, Lomborg says that these engineering projections of below-zero costs are highly unlikely. Why? Because economists think so. His discussion gives no clue that engineers might have retorts to the economists' statements (despite the fact that papers voicing some of these arguments are cited in the IPCC Working Group reports), and that engineers can point to unrealistic aspects of the economists' estimates. (An interesting news item relevant to this discussion is the recent announcement by British Petroleum thatcontrary to most economic expectationsit had reached its goal of cutting 14 percent of its carbon dioxide emissions ahead of schedule and at zero cost.)

Of course, as Lomborg well knows, the Kyoto Protocol is not intended to solve the global warming problem by itself; it is a first step that establishes a framework for further mitigation steps involving international cooperation.

Some of Lomborg's sliest comments, in his book and in his reply to our article, concern the Kyoto Protocol. He states and defends his use of a 100-year projection of the Kyoto plan, the only purpose of which is to allow him to say that it would have a trivial effect on postponing global warming. Of course, as Lomborg well knows, the Kyoto Protocol is not intended to solve the global warming problem by itself; it is a first step that establishes a framework for further mitigation steps involving international cooperation. It is disingenuous, if not dishonest, for Lomborg to pretend otherwise in his text.

Lomborg's reply about the "democratic problem" of Kyoto is similarly a smokescreen. "Since [Kyoto] is the deal offered, would it not be reasonable to discuss what is the actual outcome of the deal we are talking about?...[W]ould it not be democratically more honest to say that the decision is not Kyoto but something much more stringent?" Yet environmentalists do not say that Kyoto is anything but a first step; only Kyoto's opponents (like Lomborg) pretend that. No one yet knows what steps might be appropriate or possible afterward, because all parties seriously involved in the discussions recognize that those options (political, technological, economic and otherwise) will become more clear only over time. No one can even say how "stringent" or easy those steps may turn out to be (which is why Lomborg's dismissal of low and below-zero costs so biases his discussion). Lomborg's argument is the equivalent of saying that we sho uld not begin to fight a war unless we know and can evaluate the costs of all the as-yet-unplanned campaigns.

"The claim of 'unbalanced presentation of the natural science' clearly cannot be upheld," Lomborg writes in his reply, maintaining his "they ain't laid a glove on me" pose. That Lomborg can say this despite the evidence that he presents information about climate studies selectively and unrepresentatively is a tribute to his wishful thinking. Even in Lomborg's reply, he demonstrates that he will walk away from arguments in the natural sciences to take up more favorable economic ideas as suits his whim.


John Holdren has written his own reply to Lomborg's rebuttal, and I will not attempt to improve on it. I will say, however, that stripped of his rhetorical flourishes, Lomborg does nothing but evade the primary criticism that he is knocking down a strawman, in that he devotes much of his chapter to debunking the idea that the world is running out of energy, a position he ascribes to environmentalists but that serious environmentalists do not actually espouse. Scientific American is optimistic about the possibility of developing alternative energy sources that would reduce dependence on fossil fuels, but we do not share Lomborg's easy confidence that these alternatives will automatically appear as the economy demands without risk of serious disruption. Nor do we share his assuredness that the cost to the environment of digging deeper into the fossil fuel supplies will be minor. Our views reflect the consensus of opinions of the many researchers in energy and environmental fields with whom we have contact.


When it comes to population growth, Lomborg argues, "the number of people is not the problem." Counting the absolute numbers of people who are starving, sick, or otherwise disadvantaged is far less meaningful, in his view, than reckoning them as a percentage of the total. He summarizes the moral reasoning for this argument in a sidebar to his book, "Relative or Absolute Improvement?" (page 64), where he makes the case that given a choice between a world in which (A) 500,000 people starve out of a population of 1,000,000 or (B) 750,000 people starve out of a population of 2,000,000, one should pray for (B) because there is proportionally less misery and more happiness. (One might still imagine this would be poor consolation for the 750,000 starving, however.) A preference for smaller absolute numbers would otherwise logically lead to a further preference for (C), in which "only" 499,999 die out of 500,000. Given that the numbers of people suffering from starvation and disease are falling as a proportion of the total population and that these trends will probably continue, the world is becoming a better place.

That is the nub of Lomborg's rebuttal to overpopulation worries. The mathematical logic of his position might sound compelling. Yet in his hypothetical example, Lomborg skips one obvious possibility(D), a world in which the population stabilizes or grows slightly and the number of starving drops below 500,000. By any reckoning, that would be better option, but Lomborg ignores it, presumably because he does not consider it a realistic possibility.

What Lomborg misses about Bongaarts's point, and about the message of environmentalists in general, is that the woes of overpopulation can and should be addressed directly, with the aim of reducing both the absolute and relative numbers of the disadvantaged. Why not build on the improvements in human welfare that come from advancing technology and global wealth by slowing population growth, too?

With that in mind, let's consider Lomborg's answer to Bongaarts in more detail.

Reacting to Bongaarts's statement that "absolute growth remains close to the very high levels observed in recent decades," Lomborg writes that "today's 76 million [added to the world population] is the lowest number in the past two decades." But Lomborg's own figureshis charts of absolute and relative population growthshow this to be a precarious claim. If one goes back more than 20 years, the current absolute growth is clearly higher. Absolute growth peaked in the late 1980s, and Lomborg's graph shows that the drop-off in absolute growth will be slower than the run-up preceding that time. In fact, according to data published in 2001 by the United Nations, the projected increase for the next 50 years (from 6.1 billion to 9.3 billion) is slightly larger than the historically unprecedented increase of the past 50 years (from 2.5 billion to 6.1 billion). In the developing world, where four-fifths of the world's population lives, absolute growth in the next half century is actually projected to be slightly larger than in the past half century.

Lomborg characterizes Bongaarts's statement about "unabated" population expansion in the developing world as "misleading" because past growth "came from ever increasing numbers of people added," whereas in the future that growth will come from "ever fewer numbers being added." Bongaarts's statement is misleading only if one fixates on relative population growth, and as he has made clear, his concern is with the absolute numbersbecause from a humane perspective, every starving person counts, no matter what the larger trends may be.

Lomborg takes exceptions to Bongaarts's illustrative example of Egypt as a place where measurements of population density differ depending on whether one includes all land area or only habitable area. "Is Bongaarts obvious point at all relevant to my examples of e.g. the Netherlands being far more densely populated than India? We are never told," Lomborg sniffs.

Bongaarts has published previously on this very point. Calculating the population density in terms of the number of people per square kilometer of potentially arable land (from data circa 1990), one finds that, contrary to Lomborg, the population density in the Netherlands (385) is far lower than that of China (1,047), Bangladesh (1,188), Egypt (1,919) and India (437). (For a description of the methodology behind those figures, see "Population pressure and the food supply system in the developing world," by J. Bongaarts, in Population and Development Review, vol. 22, no. 3, pages 483-503, September 1996.)

To refute the argument that land well-suited to agriculture may be in short supply as populations expand, Lomborg uses global data from the FAO to show that the actual increase in arable land needed will be small. What his answer ignores, according to Bongaarts, is that the unused potentially arable land is distributed very unevenly around the developing world: room for agricultural expansion still exists in Latin America and Africa, but Asia has essentially run out of usable land.

Lomborg scolds Bongaarts and other environmentalists for ignoring the historical trend of falling food prices. But he does not acknowledge Bongaarts's point that environmental costsassociated with local pollution, depletion of water resources, soil erosion, and so onrarely figure into food prices. Moreover, Lomborg offers no real answer to the objection that food prices are in part kept low by subsidies from governments; he only repeats that projections by various agencies agree that prices will continue to be lower in the future (thanks to subsidies and environmental degradation that markets do not count).

"Lomborg does not deny this environmental impact but asks unhelpfully, 'What alternative do we have, with more than 6 billion people on Earth?'" Bongaarts wrote. Lomborg gives three answers in his rebuttal. The first is a reiteration of his faith in "the demographic transition," seen frequently in the past, that as countries become more wealthy, their population growth declines. In effect, Lomborg argues that these problems will correct themselves in the long run. One problem with this answer is that although it promises relief someday, it ducks the question of why more should not be done immediately to solve the current problems, especially if those immediate efforts might also help to promote the long-term economic well-being of the developing world.

Bongaarts has also sent us this note: "Up until the late 1980s many economists believed that a decline in the birth rate required conventional economic development, with rising GDP per capita, industrialization and urbanization. This view is no longer accepted, because there are now numerous examples of countries that have seen dramatic declines in birth rates while being very poor. Perhaps the most striking example is Bangladesh where the average number of children per woman has declined over the past three decades from the traditional six or more to near three today. Even larger declines in fertility have been observed in Sri Lanka and in the state of Kerala in India where fertility has actually declined to about two births per woman. All these populations are among the poorest in the world as measured by GDP per capita, but they have achieved lower birth rates by implementing family planning, health and education programs."

Lomborg's second answer to the question is irrelevant, in that he tries to show that some environmental prescriptions, carried to an extreme, might be more damaging than current practices. His hypothetical example describes what would happen if we were "to forsake fertilizer"; but serious environmentalists recommend only a reduction in synthetic fertilizer use to balance the pros and cons of the different approaches. And of course, as more is done to reduce the size of the population, less agricultural pressure is put on the environment, no matter what approach to farming is used.

Lomborg's third answer is to say that Bongaarts doesn't answer the question himself, but surely he can't believe that unless he has ignored what environmentalists have frequently said on this topic. The solution prescribed by Bongaarts (see "Population policy options in the developing world," by J. Bongaarts, in Science, vol. 263, 11 February 1994, pages 771-776) and other environmentalists is to reduce population growth through a comprehensive strategy that includes:

  • (a)Strengthening family planning programs to provide women wi th the knowledge and means to regulate their fertility. Unwanted childbearing accounts for about one quarter of all births (outside China). Around the world, a substantial proportion of those women who do not wish to get pregnant are not using contraception because it is unavailable or because they do not know enough about it. These programs can lower fertility by addressing these unmet needs.

  • (b)Emphasizing "human development," in particular education, gender equality and child health. Improvements in these areas are instrumental in reducing the desired family size. Since desired family size is still above two children in much of the developing world, population growth will continue unless fertility preferences decline further.

  • (c) Encouraging delays in childbearing. This is a relatively new but potentially effective population policy option aimed at reducing population momentum. (This momentum is the tendency of a population to keep growing even after fertility is reduced bec ause of a young population age structure, which includes large numbers of women entering their childbearing years.) One of the more desirable ways to achieve childbearing delays is by raising investments in education, especially of girls, because it is as sociated with later marriage and onset of childbearing.

When Bongaarts describes the contribution of population growth to poverty, Lomborg's reaction is that it "seems like a reasonable argumentthough unfortunately without any references." (If Lomborg would like one, he can try Population matters: Demographic change, economic growth, and poverty in the developing world, by Nancy Birdsall, Allen C. Kelley and Steven W. Sinding; Oxford University Press, 2001.) But Lomborg then seizes on Bongaarts's statement that the economic contributions of reducing fertility "can be realized only in countries with otherwise sound economic policies," and he tries to reduce Bongaarts's solution to "a subset of my point that in order to combat population growth, we need to address and reduce poverty."

By preaching that the woes associated with overpopulation will eventually take care of themselves as side effects of economic development, he discourages his readers from supporting the effective, desirable policies that could help in the short run.

Nonsense. Bongaarts is acknowledging that sound economic policies are a necessary condition for economic development; he does not claim that reducing the population would magically bring prosperity. He recommends working on reducing population growth and on reducing poverty simultaneously. Conversely, it is Lomborg who argues that economic development is a necessary and sufficient condition for reducing the population (although this disregards the examples of Ban gladesh, Sri Lanka and Kerala, noted above). Economic development alone almost certainly will reduce population growthbut only after considerably more growth has occurred and at the cost of much avoidable suffering within the developing world.

Fundamentally, this is again the problem with Lomborg's position. By preaching that the woes associated with overpopulation will eventually take care of themselves as side effects of economic development, he discourages his readers from supporting the effective, desirable policies that could help in the short run.

Of course, it may be that part of Lomborg's case is missing. Perhaps he wants to argue that economic development is impossible without a large number of poor, starving people in the develop ing world. That would certainly explain his resistance to Bongaarts's position that economic development and overpopulation can be addressed simultaneously and symbiotically. If that is Lomborg's implicit argument, he should articulate it more openly.

Lomborg questions the source of Bongaarts's assertion that "traditional urban advantage is eroding in the poorest countries." It is this: "The poverty of cities in developing regions," by Martin Brockerhoff and Ellen Brennan, in Population and Development Review, vol. 24, no. 1, pages 75-114; March 1998.

At the end of his rebuttal to Bongaarts, Lomborg criticizes this statement: "It is true that life has improved for many people in recent decades, but Lomborg does not acknowledge that this favorable trend has been brought about in part by intensive efforts by governments and the international community." Based on his response, Lomborg seems to have taken this as a criticism that he should have said "who exactly carried out the concrete improvements."

But this is a complete misreading of Bongaarts's meaninghis emphasis was not on "governments and the international community" but on "intensive efforts." Bongaarts is saying that life would have improved substantially less had so much effort not been expended towa rd combating the effects of population growth directly. Much more can be done to reduce expected rapid future growth in the developing world. This is why his concluding sentence reads, "Unfortunately, the unrelenting we-are-doing-fine tone that pervades Lomborg's book encourages complacency rather than urgency." That damning objection remains unanswered.


Lovejoy wrote that he was disconcerted to see that Lomborg's opening discussion questioned whether biodiversity is important. Lomborg raises an eyebrow over this as a peculiar reaction more befitting an environmental advocate. But as a reading of Lomborg's chapter on biodiversity reveals, that passage does not merely question whether biodiversity is importantit argues that biodiversity is not. In less than a page, Lomborg trivializes the value of a biologically diverse environment, brushing past the most serious criticism that economic analyses of biodiversity generally underestimate its value because the majority of services that the environment provides are not represented in economic markets. It sets the context for minimizing the significance of species loss that persists throughout the rest of the chapter. Lovejoy's discomfort at such a cavalier argument seems justified.

To rebut Lovejoy's criticism that he "confounds the process by which a species is judged to be extinct with the estimates and projections of extinction rates," Lomborg quotes passages in which he does make those distinctions. Yet as the substance of the rest of Lovejoy's argument shows, Lomborg does not practice what he preaches. At the least he disdains as "farcical" the difference he pretends to respect.

For example, in his rebuttal, Lomborg writes, "Yes, the Brazilian Atlantic rainforest has been cut down by about 90%--this would mean that the species-area formula would expect a loss of about half of all species." He then cites studies showing that no extinct species have been found as evidence that the species-area formula must be wrong. In practice he is ignoring the lag time between habitat destruction and species extinction.

Lomborg tries to make it seem ridiculous that species losses in that area are not more numerous by saying that the Brazilian Atlantic rain forest was cleared more than 100 years ago, in the 1800s. As Lovejoy notes, there are few field biologists surveying this region even today, so some losses that could have occurred from clearing in the 1800s might have gone completely unnoticed. Moreover, as Lovejoy recently communicated to us, most of the serious clearing of the Brazilian Atlantic rain forest has occurred since the 1950s, not the 1800s (See The Vanishing Brazilian Atlantic Rainforest, by G.A.B. Fonseca, Biological Conservation, 1985).

The Brazilian Atlantic forest is home to more threatened species as assessed by the species survival commission of the World Conservation Union (IUCN) than is any other South American ecosystem. According to Lovejoy, the forest has precisely the number of threatened endemic species one would predict given its deforestation rate based on species area curves (an example of a paper in support of this statement is "Relative Risk of extinction of passerine birds on continents and islands," by L. L. Manne, T. M. Brooks and S. L. Pimm, in Nature, vol. 399, pages 258-261; May 20, 1999).

Turning to the example of Puerto Rico and its biodiversity loss, Lomborg castigates Lovejoy for saying that he "ignores" that seven of the 60 species unique to that island were lost, when Lomborg can quote a paragraph and an endnote from his book that acknowledge this fact. It would have been more accurate for Lovejoy to have said that Lomborg dismisses those species losses as minor.

However, Lomborg does not seem to have a rebuttal for Lovejoy's more important points: that primary forest in Puerto Rico was reduced to only 10 percent, not to 1 percent; that many species of bird did not go extinct because they were not unique to Puerto Rico and could recolonize it from other islands; that the seven species that did become extinct were all from the group of only 20 bird species unique to the island; and that at least four of the surviving species have only a few members and may yet perish.

Lovejoy notes that Lomborg frequently cites Norman Myers's 1979 estimate that 40,000 species disappear every year. Lomborg reiterates the cr iticism in his rebuttal and throws in mention of Lovejoy's own 1980 estimate that 15-20% of all species might be extinct by now. Those estimates were early attempts to quantify the extinction rates; like most early attempts, they look crude in retrospect. In the two decades since those predictions were made, both those researchers have revised their estimates significantly. Unquestionably, those figures do live on in some popular writings and they should not, but they are also not the figures that serious scientists still use in their work.

Lomborg's rebuttal then turns to his preference for stating extinction rates as a percentage of all species over time rather than in absolute terms or as multiples of the background extinction rate. His justification is that "for public discussion," he thinks it would be wrong to use multiples of natural extinction rates because non-biologists would have too poor a sense of what a rate 1,000-1,500 times the natural rate would mean. This is a circumspect way of saying that the public might be worried that there is a problem, whereas an extinction rate of 0.7 percent of all species over the next 50 years sounds less alarming.

Lomborg's approach has the effect of sidestepping how significantly human activities affect the extinction rate. Biodiversity biologists' point is that humans are greatly speeding up the rate at which species are lost. Knowing the magnitude of the human influence is particularly important for policy makers. Imagine that a particularly brilliant course of environmental policies could cut the human influence on extinction in half while imposing a small economic burden. A public that knows the action would cut the extinction rate from 1,000 times the background level to only 500 times could find reasonable motivation to act. A public hearing that extinction rates would drop from 0.7 percent per 50 years to 0.35 percent might wonder why it should bother.

"Lovejoy does not actually challenge my central prediction of 0.7% species loss over the next 50 years," Lomborg writes, adding that this lack of challenge "should have been mentioned in the Scientific American piece." His book equates that figure to about 1,000-1,500 times the background extinction rate, as biologists generally reckon it. There is no reason for Lovejoy to challenge that estimate, because within a wide range of estimates, that is roughly the figure that most environmental biologists currently tend to favor --including Lovejoy, Myers and E. O. Wilson, although you would never know it from Lomborg's discussion. In fact, it is surprising that Lomborg supports this estimate, because the evidence for it is built in part on the kinds of species-area models and data that he disparages in hi s chapter, where he also insinuates that biologists would like to keep the extinction estimates inflated because there "are many grants at stake."

But if Lomborg and Lovejoy more or less agree about that extinction rate estimate, why are Lovejoy and other biologists so upset with Lomborg's argument? First, as noted previously, because he minimizes the significance of biodiversity loss. From his opening section in which he argues that biodiversity has little value, one might think it would be okay even if the extinction rate were 7.0 percent per 50 years, not 0.7 percent. (Why, in fact, is there any incentive to decrease extinction rates in Lomborg's view?)

Second, biologists criticize Lomborg because he largely waves away evidence that extinction rates might be accelerating. As Lomborg writes in his rebuttal, "Lovejoy ends by telling us that it will get much worse, as natural habitats continue to dwindle. This of course is a question of whether we believe that forests (especially tropical forests) will continue to be cut down." He then cites his evidence for thinking that in 2100 there will be at least as much forest as there is today. That argument glides past crucial points, however. One is that Lomborg's treatment of forest data ignores differences that environmental biologists regard as significant--that newly planted secondary forest is not the equivalent of old-growth primary forest. Another is that by jumping a century into the future, Lomborg can retreat to the safety of saying that these extinction problems are purely temporary, and that in the long run, things will work out. But of course, 2100 is an arbitrary date; he might just as easily have said 2150 or 2200 and it would have been no less trueand no more reassuringthat eventually extinction would cease to be a problem on its own.

Lomborg is also missing an important argument that Wilson and other environmental biologists have made about biodiversity. Let's assume that Lomborg is correct and that a century from now the human pressures on biodiversity will be diminishing (a very real possibility). Then if we do value biodiversity, the challenge is to make sure as many species as possible pass through this "bottleneck" of extinction in the coming decades. Every species that disappears between now and then is gone forever. Preserving biodiversity should therefore rise as a short-term priority, contrary to Lomborg's implications that we can afford to wait.

Asserting that acidic deposition has not been shown to be the predominant cause of any forest decline, Lomborg defends himself by citing the report of the National Acid Precipitation Assessment Program (NAPAP). Readers may not know that the NAPAP paper, which dates from 1990, was considered controversial at the time of its release, and its conclusions were hotly debated. The NAPAP's Biennial Report to Congress from 1998, updating the earlier work, is available on the web. To quote from its executive summary: "Sulfur and nitrogen deposition have caused adverse impacts on certain highly sensitive forest ecosystems in the United States. High-elevation spruce-fir forests in the eastern United States are the most sensitive. Most forest ecosystems in the East, South, and West are not currently known to be adversely impacted by sulfur and nitrogen deposition. However, if deposition levels are not reduced in areas where they are presently high, adverse effects may develop in more forests due to chronic, multiple decade exposure." An even more up-to-date reference that comments on sugar maple declines in the northeast is "Acidic deposition in the northeastern United States: Sources and inputs, ecosystems effects, and management strategies," by C. T. Driscoll et al., Bioscience, vol. 51, pages 180-198.

If only in fairness to his readers, Lomborg might originally have indicated that he was throwing out a set of data based on his own opinion.

In defense of his use of FAO records on forests, Lomborg says that this data series is the only one in existence, that he notes its uncertainties, and that some experts also use it in their work. Most researchers, faced with such uncertain evidence, would argue for conclusions based on this flawed data set with less confidence than Lomborg does.

When Lovejoy indicates Lomborg's selective citation of data concerning the huge Indonesia fires of 1997-1998, Lomborg's retort is that "I clearly do not accept the Indonesian estimate." If only in fairness to his readers, Lomborg might originally have indicated that he was throwing out a set of data based on his own opinion. He may be entitled to doubt the motives and credibility of the Indonesian paper's authors, but he shows a more forgiving attitude toward sources that support his case elsewhere in the book. In any case, his statement does nothing to deny that he is citing data selectively.

For a more complete assessment on the Indonesia fires, I offer this summary by Emily Matthews of the World Resources Institute (from her entry in the series of rebuttals to Lomborg found on "An analysis by the Singapore Centre of Remote Imaging, Sensing, and Processing using the same satellite images yielded a total burn area estimate for 1997 and 1998 of nearly 8 million hectares. In 1999, a technical team funded by the Asian Development Bank and working through the Indonesian National Development Planning Agency aggregated and analyzed all available data sources and estimated that the area burned during 1997-1998 totaled more than 9.7 million hectares, of which some 4.6 million hectares were forest. Thus, the most authoritative consensus estimate of the extent of forests burned during the Indonesian fires of 1997-1998 is more than twice the WWF estimate that is derided by Lomborg."

It is strange to read that Lomborg thinks Lovejoy fails to establish "a pattern of denial" in Lomborg's writings about deforestation, acid rain and other environmental issuesparticularly when Lomborg has previously argued in his rebuttal that forest cover will increase and that acid rain has not been a predominant cause of forest decline. Perhaps this means that Lomborg believes he is not denying those phenomena if he acknowledges that they are problems to any degree, no matter how minor.

One of the most amazing passages in Lomborg's rebuttal is his answer to Lovejoy's accusation that he took a quotation from Paul A. Colinvaux out of context to distort its meaning. Lomborg wrote that "Colinvaux admits in Scientific American that the rate [of extinction in the tropics] is 'incalculable.'" As Lovejoy correctly indicates, the original sentence of Colinvaux's article says, "As human beings lay waste to massive tracts of vegetation, an incalculable and unprecedented number of species are rapidly becoming extinct." Colinvaux was using the word "incalculable" in the sense of "countless" or "vast," and nothing in his article suggests that he was "admitting" anything by using it.

Clearly, Lomborg has misused the Colinvaux quotation, and one might assume that faced with this evidence, Lomborg would simply acknowledge the error. Instead, his rebuttal emphasizes that he was trying to make a larger point: "Here it is evident that I am trying to establish the fact that the vast extinction numbers are unsupported." Apparently, his defense is that it is okay to take someone else's remarks out of context and change their meaning as long as they make sense in the context of his own remarks.

He also writes, "Of course, Lovejoy would like me to quote that Colinvaux really does believe that the number is large, but this is a personal and unsubstantiated point." That remark might make sense if the word "incalculable" in Colinvaux's sentence had some greater factual basis and intent than all the rest, but that is not the case outside of Lomborg's imagination. Colinvaux was not even trying to quantify the extinction rate, so his calling it "in calculable" could not be meant as a statement about it.

Perhaps this seems like a disproportionate amount of evidence to bring against one tiny statement in Lomborg's book. Yet Lomborg's misuse of the quote is transparent, the kind of clumsy mistake one might find on a very poor student's homework. Instead of acknowledging it, however, Lomborg chose to defend it. To the extent that this is representative of Lomborg's scholarship, it should make anyone uneasy about the reliability of his arguments.

Lomborg asks Lovejoy to provide an example of references that could not be tracked down, and Lovejoy offers an obvious one: Lomborg's figure for the extinction rate, 0.7 percent of all species over the next 50 years. The footnoted reference is to an article by Nigel E. Stork, listed as being in "Wilson, et al. 1997:41-68." But the bibliography does not list a "Wilson, et al." The Stork article actually appears in the bibliography in a different collectionBiodiversity II, edited by Marjorie Reaka-Kudla, et al., 1997. No one unfamiliar with the literature could have figured this out easily.

Lomborg closes his rebuttal to Lovejoy by arguing that the activist efforts of environmentalists and government policies have had little to do with improvem ents in air quality. Most of the credit for better air, he says, really belongs to "improved products and technology for industry and the home." Yet this argument implies that the only way in which environmentalists have their effect is by enacting new pollution laws. By making the public and industry aware of the problems associated with a damaged environment, environmentalists help to create a demand for those more ecologically beneficial products and technologies in the public mind.

Contrary to Lomborg, Lovejoy is not trying to ascribe all environmental progress to "environmentalists," and certainly not if environmentalists are supposed to be a group somehow different from "all the scientists and politicians working hard on practical solutions to the very real pollution problems," a bizarre schism that Lomborg sets up. If anything, it is Lomborg who is trying to make "environmentalist" an epithet that he can apply to any scientist with whom he disagrees.

Back to Skepticism toward The Skeptical Environmentalist