Download pdf - Science against The Skeptical Environmentalist MISLEADING …kotliar/278/Skeptical...Measuring the Real State of the World. A “skeptical environmentalist” is cer-tainly the best

GE

RR

Y E

LLIS

Min

den

Pic

ture

s

conflict—they deserve full scrutiny. With that in mind, thebook The Skeptical Environmentalist (Cambridge UniversityPress), by Bjørn Lomborg, a statistician and political scien-tist at the University of Aarhus in Denmark, should be a wel-come audit. And yet it isn’t.

As the book’s subtitle—Measuring the Real State of theWorld—indicates, Lomborg’s intention was to reanalyze en-vironmental data so that the public might make policy deci-sions based on the truest understanding of what sciencehas determined. His conclusion, which he writes surprisedeven him, was that contrary to the gloomy predictions ofdegradation he calls “the litany,” everything is getting bet-ter. Not that all is rosy, but the future for the environmentis less dire than is supposed. Instead Lomborg accuses apessimistic and dishonest cabal of environmental groups,institutions and the media of distorting scientists’ actualfindings. (A copy of the book’s first chapter can be found atwww.lomborg.org)

The problem with Lomborg’s conclusion is that the sci-entists themselves disavow it. Many spoke to us at SCIENTIF-IC AMERICAN about their frustration at what they described

as Lomborg’s misrepresentation of their fields. His seem-ingly dispassionate outsider’s view, they told us, is oftenmarred by an incomplete use of the data or a misunder-standing of the underlying science. Even where his statisti-cal analyses are valid, his interpretations are frequentlyoff the mark—literally not seeing the state of the forestsfor the number of the trees, for example. And it is hard notto be struck by Lomborg’s presumption that he has seeninto the heart of the science more faithfully than have in-vestigators who have devoted their lives to it; it is equallycurious that he finds the same contrarian good news lurk-ing in every diverse area of environmental science.

We asked four leading experts to critique Lomborg’streatments of their areas—global warming, energy, popula-tion and biodiversity—so readers could understand whythe book provokes so much disagreement. Lomborg’s as-sessment that conditions on earth are generally improvingfor human welfare may hold some truth. The errors de-scribed here, however, show that in its purpose of describ-ing the real state of the world, the book is a failure.

MISLEADING MATH

EARTH

MISLEADING MATHabout theEARTH

Science defends itself

against The Skeptical

Environmentalist

John Rennie, EDITOR IN CHIEF

CRITICAL thinking and hard data are cornerstones ofall good science. Because environmental sciences are so keenly important to both our biological and economic survival—causes that are often seen to be in

Copyright 2001 Scientific American, Inc.Copyright 2001 Scientific American, Inc.

or three decades, I havebeen debating alternativesolutions for sustainabledevelopment with thou-sands of fellow scientistsand policy analysts—ex-changes carried out in

myriad articles and formal meetings.Despite all that, I readily confess a lin-gering frustration: uncertainties so infusethe issue of climate change that it is stillimpossible to rule out either mild or cat-astrophic outcomes, let alone provideconfident probabilities for all the claimsand counterclaims made about environ-mental problems.

Even the most credible internationalassessment body, the IntergovernmentalPanel on Climate Change (IPCC), hasrefused to attempt subjective probabilis-tic estimates of future temperatures.This has forced politicians to make theirown guesses about the likelihood of var-ious degrees of global warming. Willtemperatures in 2100 increase by 1.4degrees Celsius or by 5.8? The differ-ence means relatively adaptable changesor very damaging ones.

Against this background of frustra-tion, I began increasingly to hear that ayoung Danish statistician in a politicalscience department, Bjørn Lomborg,had applied his skills in statistics to bet-ter determine how serious environmen-tal problems are. Of course, I was anx-ious to see this highly publicized contri-bution—The Skeptical Environmentalist:

Measuring the Real State of the World.A “skeptical environmentalist” is cer-tainly the best kind, I mused, becauseuncertainties are so endemic in thesecomplex problems that suffer from miss-ing data, incomplete theory and nonlin-ear interactions. But the “real state of theworld”—that is a high bar to set, giventhe large range of plausible outcomes.

And who is Lomborg, I wondered,and why haven’t I come across him atany of the meetings where the usual sus-pects debate costs, benefits, extinctionrates, carrying capacity or cloud feed-back? I couldn’t recall reading any sci-entific or policy contributions from himeither. But there was this massive 515-page tome with a whopping 2,930 end-notes to wade through. On page xx ofhis preface, Lomborg admits, “I am notmyself an expert as regards environmen-tal problems”—truer words are notfound in the rest of the book, as I’ll soonillustrate. I will report primarily on thethick global warming chapter and its600-plus endnotes. That kind of dead-weight of detail alone conjures at leastthe trappings of comprehensive andcareful scholarship. So how does the re-ality of the text hold up to the pretense?I’m sure you can already guess, but let megive some examples to make clear what Ilearned by reading.

The climate chapter makes four ba-sic arguments:

Climate science is very uncertain,but nonetheless the real state of the sci-

ence is that the sensitivity of the climateto carbon dioxide will turn out to be atthe low end of the IPCC uncertaintyrange—which is for a warming of 1.5 to4.5 degrees C if carbon dioxide were todouble and be held fixed over time.

Emissions scenarios, according tothe IPCC, fall into six “equally sound”alternative paths. These paths span adoubling in carbon dioxide concentra-tions in 2100 up to more than triplingand well beyond tripling in the 22ndcentury. Lomborg, however, dismissesall but the lowest of the scenarios:“Temperatures will increase much lessthan the maximum estimates fromIPCC—it is likely that the temperaturewill be at or below the B1 estimate [thelowest emissions scenario] (less than 2°C in 2100) and the temperature will cer-tainly not increase even further into thetwenty-second century.”

Cost-benefit calculations show thatalthough the benefits of avoiding cli-mate change could be substantial ($5trillion is the single figure Lomborgcites), this is not worth the cost to theeconomy of trying to constrain fossil-fuel emissions (a $3-trillion to $33-tril-lion range he pulls from the economicsliterature). Asymmetrically, no range isgiven for the climate damages.

The Kyoto Protocol, which caps in-dustrialized countries’ output of green-house gases, is too expensive. It wouldreduce warming in 2100 by only a fewtenths of a degree—“putting off the tem-perature increase just six years.” Thisnumber, though, is based on a straw-man policy that nobody has seriouslyproposed: Lomborg extrapolates theKyoto Protocol, which is applicableonly up to 2012, as the world’s sole cli-mate policy for another nine decades.

Before providing specifics of why Ibelieve each of these assertions is fatally R

ICH

ARD

HAM

ILTO

N S

MIT

H C

orb

is

Stephen Schneider

GLOBAL WARMING:NEGLECTING THE COMPLEXITIES

F

Copyright 2001 Scientific American, Inc.

flawed, I should say something aboutLomborg’s methods. First, most of hisnearly 3,000 citations are to secondaryliterature and media articles. Moreover,even when cited, the peer-reviewed arti-cles come elliptically from those studiesthat support his rosy view that only thelow end of the uncertainty ranges will beplausible. IPCC authors, in contrast,

were subjected to three rounds of reviewby hundreds of outside experts. Theydidn’t have the luxury of reporting pri-marily from the part of the communitythat agrees with their individual views.

Second, it is ironic that in a popularbook by a statistician one can’t find aclear discussion of the distinction amongdifferent types of probabilities, such asfrequentist and Bayesian (that is, “ob-jective” and “subjective”). He uses theword “plausible” often, but, curiouslyfor a statistician, he never attaches anyprobability to what is “plausible.” TheThird Assessment Report of the IPCC,on the other hand, explicitly confrontedthe need to quantify all confidenceterms. Working Group I, for example,gave the term “likely” a 66 to 90 per-cent chance of occurring. Although theIPCC gives a wide range for most of itsprojections, Lomborg generally dismiss-es these ranges, focusing on the least se-rious outcomes. Not so much as oneprobability is offered for the chance of adangerous outcome, yet he makes a firmassertion that climate “will certainly”not go beyond 2 degrees C warming inthe 22nd century—a conclusion at vari-

ance with the IPCC, other national cli-mate assessments and most recent stud-ies in the field of climate science.

Now let us look in more detail atthe four major arguments he makes inthis chapter.

Climate science. A typical exampleof Lomborg’s method is his paraphraseof a secondary source in reporting a

1989 Hadley Center paper in the jour-nal Nature in which the researchers makemodifications to their climate model:“The programmers then improved thecloud parameterizations in two places,and the model reacted by reducing itstemperature estimate from 5.2° C to1.9° C.” Had this been first-rate scholar-ship, Lomborg would have consultedthe original article, in which the conclud-ing sentence of the first paragraph pre-sents the authors’ caveat: “Note that al-though the revised cloud scheme is moredetailed it is not necessarily more accu-rate than the less sophisticated scheme.”

In a similar vein, he cites Richard S.Lindzen’s controversial stabilizing feed-back, or “iris effect,” as evidence thatthe IPCC climate sensitivity rangeshould be reduced by a factor of almostthree. He fails either to understand thismechanism or to tell us that it is basedon only a few years of data in a smallpart of one ocean. Extrapolating thissmall sample of data to the entire globeis like extrapolating the strong desta-bilizing feedback over midcontinentallandmasses as snow melts during thespring—such an inappropriate projec-

tion would likely increase estimates ofclimate sensitivity by a factor of several.

As a final example, he quotes a con-troversial hypothesis from Danish cloudphysicists that solar magnetic eventsmodulate cosmic rays and produce “aclear connection between global low-level cloud cover and incoming cosmicradiation.” The Danish researchers use

this hypothesis to support an alternativeto carbon dioxide for explaining recentclimate change. Lomborg fails to dis-cuss—and I haven’t seen it treated bythe authors of that speculative theory ei-ther—what such purported changes tothis cloud cover have done to the radia-tive balance of the earth. Increasingclouds, it has been well known since pa-pers by Syukuro Manabe and RichardT. Wetherald in 1967 and myself in1972, can warm or cool the atmospheredepending on the height of the cloudtops, the reflectivity of the underlyingsurface, the season and the latitude. Thereason the IPCC discounts this theory isthat its advocates have not demonstrat-ed any radiative forcing sufficient tomatch that of much more parsimonioustheories, such as anthropogenic forcing.

Emissions scenarios. Lomborg as-serts that over the next several decadesnew, improved solar machines and oth-er renewable technologies will crowdfossil fuels off the market. This will bedone so efficiently that the IPCC scenar-ios vastly overestimate the chance formajor increases in carbon dioxide. HowI wish this would turn out to be true!

Lomborg admits, “I am not myself an expert as regards ENVIRONMENTAL PROBLEMS”—truer words

are not found in the rest of the book.


64 S C I E N T I F I C A M E R I C A N J A N U A R Y 2 0 0 2

But wishes aren’t analysis. One study iscited; ignored is the huge body of eco-nomics work he later accepts to estimatea range of costs if we were to implementemissions controls. In fact, most of theseeconomists strongly believe high emis-sions are quite likely: they usually proj-ect carbon dioxide doubling to tripling

(or more) as “optimal” economic poli-cy. I have attacked this literature forfailing to point out that climate policiesthat raise the price of conventional fuelsspur investments in alternative energysystems. But such incentives need poli-cies first—and Lomborg opposes thosevery policies. No credible analyst canjust assert that a fossil-fuel-intensive sce-nario is not plausible—and, typically, hegives no probability that it might occur.

Cost-benefit calculations. Lomborg’smost egregious distortions and poorestanalyses are his citations of cost-benefitcalculations. First, he chides the govern-ments that modified the penultimatedraft of the report from IPCC’s Work-ing Group II. These modifications down-graded the significance of economic stud-ies that aggregate climate change dam-ages. Lomborg says: “A political decision

stopped IPCC from looking at the totalcost-benefit of global warming.” (As anaside, I should mention that it is strangehe chose to cite the penultimate and pre-approval draft report in this case butdidn’t mention the very first item in theapproved summary—that recent tem-perature trends have caused a dis-

cernible effect on plants and animals.Even more puzzling is his failure to dis-cuss ecological impacts in general, fo-cusing instead on health and agriculture,sectors he thinks won’t be much harmedby climate change of the minusculeamount he predicts.)

The government representativesdowngraded aggregate cost-benefit stud-ies for a reason: these studies fail to con-sider so many categories of damagesheld to be important by political leadersas to render them just a guideline on mar-ket-sector transactions, not the “totalcost-benefit” analysis Lomborg wants. Atotal analysis would have to include thevalue of species lost, crucial ecosystemservices degraded, inequity created bythe poor being hurt more than the rich(which Lomborg does acknowledge),quality of life reduced (for example, a

rise in sea level driving small-island in-habitants from traditional homelands),and likely changes to climatic extremesand variability. Then again, Lomborgcites only one value for climate dam-ages—$5 trillion—even though the sameeconomics papers he refers to for costsof climate policy generally acknowledge

that climate damages can vary frombenefits up to catastrophic losses.

It is precisely because the responsiblescientific community cannot rule outsuch catastrophic outcomes at a highlevel of confidence that climate mitiga-tion polices are seriously proposed. Andto give one number—rather than a broadrange—for avoided climate damages de-fies explanation, especially when he doesgive a range for climate policy costs. Thisrange, however, is based on the eco-nomics literature but ignores the find-ings of engineers. Engineers dispute theeconomists’ typical estimates becausethe economists fail to take into accountpreexisting market imperfections suchas energy-inefficient machines, housesand processes. These engineering stud-ies, including a famous one by five U.S.Department of Energy laboratories—

hardly environmental radicals—suggestthat climate policies that provide incen-tives to replace inefficient equipmentwith more efficient state-of-the-art prod-ucts could actually reduce some emis-sions at below-zero costs.

The Kyoto Protocol. Lomborg’screation of a 100-year regime for adecade-long protocol is a distortion ofthe climate policy process. Every IPCCreport has noted that carbon dioxideemissions need to be cut by more than50 percent below most baseline projec-tions to avoid large increases in concen-tration in the late 21st and 22nd cen-turies. Most analysts know “Kyoto ex-tended” can’t make such large cuts andthat both developed and developing na- JI

M B

RAN

DE

NB

UR

G M

ind

en P

ictu

res

It is precisely because the responsible scientific community cannot rule out CATASTROPHIC OUTCOMES that

climate mitigation policies are seriously proposed.


tions will have to fashion cooperativeand cost-effective solutions over time.This will take a great deal of learning-by-doing: international cooperation is not acommon experience. Kyoto is a startingpoint. And yet Lomborg, with his cre-ation of a straw-man 100-year projec-tion, would squash even this first step.

So what then is “the real state of theworld”? Clearly, it isn’t knowable intraditional statistical terms, even thoughsubjective estimates can be responsiblyoffered. The ranges presented by theIPCC in its peer-reviewed reports givethe best snapshot of the real state of cli-mate change: we could be lucky and seea mild effect or unlucky and get the cat-astrophic outcomes. The IPCC framesthe issue as a risk-management decisionabout hedging. It is not the everything-will-turn-out-fine affair that Lomborgwould have us believe.

For such an interdisciplinary topic,the publisher would have been wise toask natural scientists as well as social sci-entists to review the manuscript, whichwas published by the social science sideof the house. It’s not surprising that thereviewers failed to spot Lomborg’s un-balanced presentation of the natural sci-ence, given the complexity of the manyintertwining fields. But that the naturalscientists weren’t asked is a serious omis-sion for a respectable publisher such asCambridge University Press.

Unfortunately, angry reviews suchas this one will be the result. Worse still,many laypeople and policymakers won’tsee the reviews and could well be trickedinto thinking thousands of citations andhundreds of pages constitute balancedscholarship. A better rule of thumb is tosee who talks in ranges and subjectiveprobabilities and to beware of the mythbusters and “truth tellers.”

Stephen Schneider, professor in thedepartment of biological sciences andsenior fellow at the Institute forInternational Studies at StanfordUniversity, is editor of Climatic Changeand the Encyclopedia of Climate andWeather and lead author of severalIPCC chapters and the IPCC guidancepaper on uncertainties.

w w w . s c i a m . c o m S C I E N T I F I C A M E R I C A N 65

omborg’s chapter on energy cov-ers a scant 19 pages. It isdevoted almost entirely toattacking the belief thatthe world is running outof energy, a belief thatLomborg appears to re-

gard as part of the “environmentallitany” but that few if any environmen-talists actually hold. What environmen-talists mainly say on this topic is not thatwe are running out of energy but that weare running out of environment—that is,running out of the capacity of air, water,soil and biota to absorb, without intoler-able consequences for human well-being,the effects of energy extraction, trans-port, transformation and use. They alsoargue that we are running out of theability to manage other risks of energysupply, such as the political and eco-nomic dangers of overdependence onMiddle East oil and the risk that nuclearenergy systems will leak weapons mate-rials and expertise into the hands of pro-liferation-prone nations or terrorists.

That “the energy problem” is notprimarily a matter of depletion of re-sources in any global sense but rather ofenvironmental impacts and sociopoliti-cal risks—and, potentially, of risingmonetary costs for energy when its en-vironmental and sociopolitical hazardsare adequately internalized and insuredagainst—has in fact been the main-stream environmentalist position fordecades. It was, for example, the posi-tion I elucidated in the 1971 Sierra Club“Battlebook” Energy (co-authored withPhilip Herrera, then the environmenteditor for Time). It was also the positionelaborated on by the Energy Policy Proj-ect of the Ford Foundation in the pio-neering 1974 report A Time to Choose;by Amory Lovins in his influential 1976Foreign Affairs article “Energy Strategy:

The Road Not Taken”; by Paul R. andAnne H. Ehrlich and me in our 1977college textbook Ecoscience; and so on.

So whom is Lomborg so resounding-ly refuting with his treatise on the abun-dance of world energy resources? Itwould seem that his targets are pundits(such as the correspondents for E maga-zine and CNN cited at the opening ofthis chapter) and professional analysts(although only a few of these are cited,and those very selectively) who have ar-gued not that the world is running outof energy altogether but only that itmight be running out of cheap oil. Lom-borg’s dismissive rhetoric notwithstand-ing, this is not a silly question, nor onewith an easy answer.

Oil is the most versatile and currentlythe most valuable of the conventionalfossil fuels that have long provided thebulk of civilization’s energy, and it re-mains today the largest contributor toworld energy supply (accounting fornearly the whole of energy used fortransport, besides other roles). But therecoverable conventional resources ofoil are believed (on substantial evi-dence) to be far smaller than those ofcoal and probably also smaller thanthose of natural gas; the bulk of theseresources appears to lie in the politicallyvolatile Middle East; much of the restlies offshore and in other difficult or en-vironmentally fragile locations; and it islikely that the most abundant potentialreplacements for conventional oil willbe more expensive than oil has been.For all these reasons, concerns aboutdeclining availability and rising priceshave long been more salient for oil thanfor the other fossil fuels. There is, ac-cordingly, a serious technical literature(produced mainly by geologists andeconomists) exploring the questions ofwhen world oil production will peak

John P. Holdren

ENERGY: ASKING THE WRONG QUESTION

L


and begin to decline and what the priceof oil might be in 2010, 2030 or 2050,with considerable disagreement amonginformed professionals on the answers.

Lomborg gets right the basic pointthat the dominance of oil in the world en-ergy market will end not because no oil isleft in the ground but because other ener-

gy sources have become more attractiverelative to oil. But he seems not to rec-ognize that the transition from oil to oth-er sources will not necessarily be smoothor occur at prices as low as those enjoyedby oil consumers today. Indeed, whileridiculing the position that the world’sheavy oil dependence may again proveproblematic in our lifetimes, he showsno sign of understanding (or no interestin communicating) why there is real de-bate among serious people about this.

Lomborg does not so much as offerhis readers a clear explanation of thedistinction—crucial to understandingarguments about depletion—between“proved reserves” (referring to materialthat has already been found and is ex-ploitable at a profit at today’s prices, us-ing today’s technologies) and “remain-ing ultimately recoverable resources”(which incorporate estimates of addi-tional material exploitable with today’stechnology at today’s prices but still tobe found, as well as material both al-

ready found and still to be found thatwill be exploitable with future technolo-gies at potentially higher future prices).And, while noting that most of theworld’s oil reserves lie in the Middle East(and failing to note, having not even in-troduced the concept, that a still largershare of remaining ultimately recover-

able resources is thought to lie there), heplacidly informs us that it is “imperativefor our future energy supply that this re-gion remains reasonably peaceful,” as ifthat observation did not undermine anybasis for complacency. (At this juncture,one of his 2,930 footnotes helpfully addsthat this peace imperative for the MiddleEast was “one of the background rea-sons for the Gulf War”!)

Lomborg’s treatment of energy re-sources other than oil is not much bet-ter. He is correct in his basic propositionthat resources of coal, oil shale, nuclearfuels and renewable energy are immense(which few environmentalists—and nowell-informed ones—dispute). But hishandling of the technical, economic andenvironmental factors that will governthe circumstances and quantities inwhich these resources might actually beused is superficial, muddled and oftenplain wrong. His mistakes include ap-parent misreadings or misunderstand-ings of statistical data—in other words,

just the kinds of errors he claims arepervasive in the writings of environmen-talists—as well as other elementaryblunders of quantitative manipulationand presentation that no self-respectingstatistician ought to commit.

He tells us correctly, for example,that the world has huge resources of coal,

but in observing that “it is presumed thatthere is sufficient coal for well beyondthe next 1,500 years” he says nothingabout the rate of coal use for which thisconclusion might obtain. Concerning theenvironmental questions that increasedreliance on coal would raise, he writesthe following: “Typically, coal pollutesquite a lot, but in developed economiesswitches to low-sulfur coal, scrubbersand other air-pollution control deviceshave today removed the vast part of sul-fur dioxide and nitrogen dioxide emis-sions.” To the contrary, data readilyavailable on the Web in the Environ-mental Protection Agency report Na-tional Air Pollutant Emission Trends1900–1998 reveal that U.S. emissionsof nitrogen oxides from coal-burningelectric power plants were 6.1 millionshort tons in 1980 and 5.4 millionshort tons in 1998. Emissions of sulfurdioxide from U.S. coal-burning powerplants were 16.1 million short tons in1980 and 12.4 million short tons in

LEST

ER

LE

FKO

WIT

Z FP

GIn

tern

ati

ona

l

What environmentalists mainly say on this topic is not that we are running out of energy but

that we are RUNNING OUT OF ENVIRONMENT.


1998. These are moderate reductions,welcome but hardly the “vast part” ofthe emissions.

Concerning nuclear energy, Lom-borg tells us that it “constitutes 6 per-cent of global energy production and 20percent in the countries that have nu-clear power.” The first figure is right,the second seriously wrong. Nuclear en-ergy provides a bit less than 10 percentof the primary energy supply in thecountries that use this energy source. (Itappears that Lomborg has confusedcontributions to the electricity sectorwith contributions to primary energysupply.) After a muddled discussion ofthe relation between uranium-resourceestimates and breeding (which omits al-together the potentially decisive issue ofthe usability of uranium from seawater),he then barely notes in passing thatbreeder reactors “produce large amountsof plutonium that can be used for nu-clear weapons production, thus addingto the security concerns.” He shouldhave added that this problem is so sig-nificant that it may preclude use of thebreeding approach altogether, unlesswe develop technologies that makebreeding much less susceptible to diver-sion of the plutonium while not makingthis approach even more uneconomicthan it is today.

Lomborg has some generally sensiblethings to say about the large contribu-tions that are possible from increased en-ergy end-use efficiency and from renew-able energy—on these topics he seems, tohis credit, to be more a contributor tothe “environmental litany” than a criticof it. But on these subjects as on the oth-ers, his treatment is superficial, unevenand marred by numerous errors and in-felicities. For example, he persistentlypresents numbers to two- and three-fig-ure precision for quantities that cannotbe known to such accuracy: “43 percentof American energy use is wasted”; “thecosts of carbon dioxide” emissions are“0.64 cents per kWh”; plant photosyn-thesis is “1,260 EJ” annually. He makesclaims, based on single citations andwithout elaboration, that are far fromrepresentative of the literature: “Weknow today that it is possible to produce

safe cars getting more than 50–100 kmper liter (120–240 mpg).” (How bigwould these cars be, and powered how?)He bungles terminology: “Energy can bestored in hydrogen by catalyzing water.”(He must mean “by electrolyzing water”or “by catalytic thermochemical decom-position of water.”) And he propagates avariety of conceptual confusions, such asthe idea that grid-connected wind powerrequires “a sizeable excess capacity” inthe windmills because these alone “needto be able to meet peak demand.”

Of course, much of what is mostproblematic in the global energy pictureis covered by Lomborg not in his energychapter but in those that deal with airpollution, acid rain, water pollution andglobal warming. The last is devastating-ly critiqued by Stephen Schneider onpage 62. There is no space to deal with

the other energy-related chapters; suf-fice it to say that I found their level ofsuperficiality, selectivity and misunder-standing roughly consistent with that ofthe energy chapter reviewed here. Thisis a shame. Lomborg is giving skepti-cism—and statisticians—a bad name.

John P. Holdren is the Teresa and JohnHeinz Professor of EnvironmentalPolicy at the John F. Kennedy School ofGovernment, as well as professor ofenvironmental science and public policyin the department of earth and planetarysciences, at Harvard University. From1973 to 1996 he co-led the interdisci-plinary graduate program in energy andresources at the University of California,Berkeley. He is a member of theNational Academy of Sciences and theNational Academy of Engineering.


round the world, countries areexperiencing unprecedent-ed demographic change.The best-known exampleis an enormous expan-sion in human numbers,but other important de-

mographic trends also affect humanwelfare. People are living longer andhealthier lives, women are bearing fewerchildren, increasing numbers of migrantsare moving to cities and to other coun-tries in search of a better life, and popu-lations are aging. Lomborg’s unbalancedpresentation of some of these trends andtheir influences emphasizes the goodnews and neglects the bad. Environmen-talists who predicted widespread famineand blamed rapid population growthfor many of the world’s environmental,economic and social problems overstat-ed their cases. But Lomborg’s view that“the number of people is not the prob-lem” is simply wrong.

His selective use of statistics givesthe reader the impression that the popu-lation problem is largely behind us. Theglobal population growth rate has in-deed declined slowly, but absolutegrowth remains close to the very highlevels observed in recent decades, be-cause the population base keeps expand-ing. World population today stands atsix billion, three billion more than in1960. According to U.N. projections,another three billion will likely be addedby 2050, and population size will even-tually reach about 10 billion.

Any discussion of global trends ismisleading without taking account of theenormous contrasts among world re-gions. Today’s poorest nations in Africa,Asia and Latin America have rapidlygrowing and young populations, where-as in the technologically advanced andricher nations in Europe, North Ameri-ca and Japan, growth is near zero (or, insome cases, even negative), and popula-

John Bongaarts

POPULATION:IGNORING ITS IMPACT

A



tions are aging quickly. As a conse-quence, nearly all future global growthwill be concentrated in the developingcountries, where four fifths of the world’spopulation lives. The projected rise inpopulation in the developing world be-tween 2000 and 2025 (from 4.87 to6.72 billion) is actually just as large as

the record-breaking increase in the pastquarter of a century. The historicallyunprecedented population expansion inthe poorest parts of the world continueslargely unabated.

Past population growth has led tohigh population densities in many coun-tries. Lomborg dismisses concerns aboutthis issue based on a simplistic and mis-leading calculation of density as the ratioof people to all land. Clearly, a moreuseful and accurate indicator of densitywould be based on the land that remainsafter excluding areas unsuited for hu-man habitation or agriculture, such asdeserts and inaccessible mountains. For

example, according to his simple calcula-tion, the population density of Egyptequals a manageable 68 persons persquare kilometer, but if the unirrigatedEgyptian deserts are excluded, density isan extraordinary 2,000 per square kilo-meter. It is therefore not surprising thatEgypt needs to import a large proportion

of its food supply. Measured properly,population densities have reached ex-tremely high levels, particularly in largecountries in Asia and the Middle East.

Why does this matter? The effect ofpopulation trends on human welfarehas been debated for centuries. Whenthe modern expansion of human num-bers began in the late 18th century,Thomas Robert Malthus argued thatpopulation growth would be limited byfood shortages. Lomborg and othertechnological optimists correctly notethat world population has expandedmuch more rapidly than Malthus envi-sioned, growing from one billion to six

billion over the past two centuries. Anddiets have improved. Moreover, thetechnological optimists are probablycorrect in claiming that overall worldfood production can be increased sub-stantially over the next few decades. Av-erage current crop yields are still belowthe levels achieved in the most produc-

tive countries, and some countries stillhave unused potential arable land (al-though much of this is forested).

Agricultural expansion, however, willbe costly, especially if global food pro-duction has to rise twofold or even three-fold to accommodate the demand forbetter diets from several billion morepeople. The land now used for agricul-ture is generally of better quality thanunused, potentially cultivable land. Simi-larly, existing irrigation systems havebeen built on the most favorable sites.And water is increasingly in short supplyin many countries as the competition forthat resource among households, indus-try and agriculture intensifies. Conse-quently, each new increase in food pro-duction is becoming more expensive toobtain. This is especially true if one con-siders environmental costs not reflectedin the price of agricultural products.

Lomborg’s view that the productionof more food is a nonissue rests heavilyon the fact that world food prices are lowand have declined over time. But this evi-dence is flawed. Massive governmentalsubsidies to farmers, particularly in thedeveloped countries, keep food prices ar-tificially low. Although technological de-velopments have reduced prices, withoutthese massive subsidies, world foodprices would certainly be higher.

The environmental cost of whatPaul R. and Anne H. Ehrlich describe as“turning the earth into a giant humanfeedlot” could be severe. A large expan-sion of agriculture to provide growingpopulations with improved diets is like-ly to lead to further deforestation, lossof species, soil erosion, and pollution D

IA M

AX F

PG

Inte

rna

tion

al

The unprecedented POPULATION EXPANSION in thepoorest parts of the world continues largely unabated.


from pesticides and fertilizer runoff asfarming intensifies and new land isbrought into production. Reducing thisenvironmental impact is possible butcostly and would obviously be easier ifpopulation growth were slower. Lom-borg does not deny this environmentalimpact but asks unhelpfully, “What al-ternative do we have, with more than 6billion people on Earth?”

Lomborg correctly notes that pover-ty is the main cause of hunger and mal-nutrition, but he neglects the contribu-tion of population growth to poverty.This effect operates through two distinctmechanisms. First, rapid populationgrowth leads to a young population, onein which as much as half is below theage of entry into the labor force. Theseyoung people have to be fed, housed,clothed and educated, but they are notproductive, thus constraining the econo-my. Second, rapid population growthcreates a huge demand for new jobs. Alarge number of applicants for a limitednumber of jobs exerts downward pres-sure on wages, contributing to povertyand inequality. Unemployment is wide-spread, and often workers in poor coun-tries earn wages near the subsistence lev-el. Both of these adverse economic effectsare reversible by reducing birth rates.With lower birth rates, schools becomeless crowded, the ratio of dependents toworkers declines as does the growth inthe number of job seekers. These benefi-cial demographic effects contributed tothe economic “miracles” of several EastAsian countries. Of course, such dramat-ic results are by no means assured andcan be realized only in countries withotherwise sound economic policies.

Lomborg approvingly notes the hugeongoing migration from villages to citiesin the developing world. This has beenconsidered a welcome development, be-cause urban dwellers generally havehigher standards of living than villagers.Because the flow of migrants is now solarge, however, it tends to overwhelmthe absorptive capacity of cities, andmany migrants end up living in appallingconditions in slums. The traditional ur-ban advantage is eroding in the poorestcountries, and the health conditions in

slums are often as adverse as in rural ar-eas. This points to another burden ofrapid population growth: the inability ofgovernments to cope with large additionsof new people. In many developing coun-tries, investments in education, healthservices and infrastructure are not keep-ing up with population growth.

It is true that life has improved formany people in recent decades, butLomborg does not acknowledge that thisfavorable trend has been brought aboutin part by intensive efforts by govern-ments and the international community.Investments in developing and distribut-ing “green revolution” technology havereduced hunger, public health campaignshave cut death rates, and family-plan-ning programs have lowered birth rates.Despite this progress, some 800 millionpeople are still malnourished, and 1.2billion live in abject poverty. This veryserious situation calls for more effectiveremedial action. Lomborg asks the de-veloped nations to fulfill their U.N.pledge to donate 0.7 percent of theirGNPs to assist the developing world, but

few countries have met this goal, and therichest nation on earth, the U.S., is oneof the stingiest, giving just 0.1 percent ofits GNP. The trend in overseas develop-ment assistance from the developed tothe developing world is down, not up.Unfortunately, the unrelenting we-are-doing-fine tone that pervades Lomborg’sbook encourages complacency ratherthan urgency.

Population is not the main cause ofthe world’s social, economic and envi-ronmental problems, but it contributessubstantially to many of them. If popula-tion had grown less rapidly in the past,we would be better off now. And if fu-ture growth can be slowed, future gener-ations will be better off.

John Bongaarts is vice president of thePolicy Research Division of the Popula-tion Council in New York City. From1998 to 2000 he chaired the Panel onPopulation Projections of the NationalAcademy of Sciences, National ResearchCouncil. He is a member of the RoyalDutch Academy of Sciences.


iologists are trained tohave a healthy respectfor statistics and statisti-cians. It was disconcert-ing, therefore, to findthat before even examin-ing the extinction prob-

lem—and the numbers invoked todemonstrate that it is or is not a prob-lem—Lomborg begins the chapter onbiodiversity with a section questioningwhether biodiversity is important. Inless than a page, he discounts its valueboth as the library for the life sciencesand as provider of ecosystem services (inpart because of a general absence ofmarkets for these services).

When he finally gets to extinction, he

totally confounds the process by whicha species is judged to be extinct with theestimates and projections of extinctionrates. Highly conservative rules holdthat to be declared officially extinct, notonly does a species have to be known toscience, it has to be observed going toextinction (as in the case of the passen-ger pigeon, the last individual of whichperished in the Cincinnati Zoo in 1914).Or, in the absence of direct observation,it must not have been seen in nature for50 years.

Projections of extinction rates, onthe other hand, are generally based onthe long-established relation betweenspecies number and area (which dates to1921, not to the 1960s, as Lomborg

B

Thomas Lovejoy

BIODIVERSITY: DISMISSING SCIENTIFIC PROCESS


maintains, and which demonstrates therate at which species number increaseswith increase in area). Researchers thenproject what the reduction in a naturalhabitat will mean in terms of speciesloss. The disappearance of a species isnot necessarily instantaneous, and thussome species that survive the initial re-duction of the habitat are essentially“living dead”—they are not able to sur-vive over the long term. The loss ofspecies from habitat remnants is a wide-ly documented phenomenon—in con-trast to Lomborg’s inclusion of an out-of-date assertion that no credible at-tempt has been made to pin down theunderlying scientific assumptions.

As a consequence, a seemingly majorcontradiction that Lomborg then offersis no contradiction at all: the reductionof the Brazilian Atlantic Forest forma-tion to something on the order of 10percent of its original extent and thelack of large numbers of recorded ex-tinctions. First, this is a region with veryfew field biologists to record eitherspecies or their extinction. Second, thereis abundant evidence that if the Atlanticforest remains as reduced and fragment-ed as it is, it will lose a sizable fractionof the species that at the moment areable to hang on.

In another supposed example of spe-cies surviving habitat loss, he notes that

few species went extinct when the easternforests of the U.S. were reduced to 1 to 2percent of their original area. But onlythe old-growth forests shrank that much;total forest cover never fell below rough-ly 50 percent—allowing much biodiversi-ty to survive as forest returned to an evengreater area. Consequently, the smallnumber of bird extinctions does not con-tradict what species-area considerationspredict but instead confirms them.

In presenting an analysis for PuertoRico, Lomborg again cites apparently

contradictory evidence that although 99percent of the primary forest was lost,the island ended up with more birdsthan it supported before deforestation.First of all, total forest cover was never sodramatically reduced. More significant,he ignores that seven of the 60 speciesunique to Puerto Rico were lost, and theadditional species are not only invasivesfrom other parts of the world but live ina wide variety of habitats. He complete-ly misses the point that the world’s birdfauna was reduced by seven species.

Lomborg takes particular exceptionto projections of massive extinction thatstarted with Norman Myers’s 1979 esti-mate that 40,000 species are being lostfrom the globe every year. There is somejustification for this objection: Myers did

not specify the method of arriving at hisestimate. Nevertheless, he deserves cred-it for being the first to say that the num-ber was large and for doing so at a timewhen it was difficult to make more accu-rate calculations. Current estimates areusually given in terms of the increasesover normal extinction rates, which ispreferable in that it is not necessary toassume a figure for the total number ofspecies on the earth. That science doesnot know the total number of speciesdoes not prevent an estimation of extinc-

tion rates. Lom-borg cynically dis-misses the use ofmultiples of normal

rates as being donebecause it sounds more

“ominous” rather thanrecognizing the al-tered approach as animprovement in thescience.

Estimates of pres-ent extinction rates

range from 100 to1,000 times normal, with

most estimates at 1,000. The percent ofbird (12), mammal (18), fish (5) andflowering plant (8) species threatenedwith extinction is consistent with thatestimate. And the rates are certain torise—and to do so exponentially—asnatural habitats continue to dwindle.

The consideration of acid rain in aseparate chapter is equally poorly re-searched and presented. Indeed, the re-search is so shallow that almost no cita-tion from the peer-reviewed literatureappears. Lomborg asserts that big-citypollution has nothing to do with acidrain, when it is fact that nitrogen com-pounds (NOx) from traffic are a majorsource. His reference to a study showingthat acid rain had no effect on theseedlings of three tree species neglects to

mention that the study did not includeconifer species such as red spruce, whichare very sensitive. There is no acknowl-edgment of the delayed effects from acidrain leaching soil nutrients, particularlykey cations. He confounds tree damagefrom air pollution 30 to 60 years agowith subsequent acid rain damage andmakes an Alice-in-Wonderland statementthat the only reason we worry about fo-liage loss is “because we have startedmonitoring this loss.” It is simply untruethat “there is no case of forest decline in


FRAN

S LA

NTI

NG

Min

den

Pic

ture

s (a

bov

e a

nd

op

pos

ite

pa

ge)

Things improve because of the efforts ofenvironmentalists to FLAG A PARTICULAR PROBLEM,

investigate it and suggest policies to remedy it.


which acidic deposition is known to be apredominant cause.” Two clear-cut ex-amples are red spruce in the Adiron-dacks and sugar maple in Pennsylvania.

The chapter on forests also suffersfrom superficial research and selectiveuse of numbers. Lomborg starts by dis-playing Food and Agriculture Organi-zation (FAO) data from 1948 to 2000.The FAO began by just reporting sumsof “official data” furnished by govern-ments (such data are notoriously un-even in quality and frequently overesti-mate forest stocks). Subsequently, theFAO adopted so many different defini-tions and methods that any statisticianshould know they could not be used fora valid time series.

Lomborg’s discussion of the great firein Indonesia in 1997 is still another in-stance of misleading readers with selec-tive information. Yes, the WWF (WorldWide Fund for Nature) first estimatedthe amount of forest burned at two mil-lion hectares, and Indonesia counteredwith official estimates of 165,000 to219,000 hectares. But Lomborg fails tomention that the latter were not in theleast credible and that in 1999 the In-donesian government and donor agen-cies, including the World Bank, signedoff on a report that the real number was4.6 million hectares.

From the very outset—his introduc-tory chapter—Lomborg confuses forestsand tree plantations. In criticizing aWWF estimate of loss of “naturalwealth,” he implies that the only valueof forests is harvestable trees. That isanalogous to valuing computer chipsonly for their silicon content. In fact, themetric the WWF used includes naturalforests (because of their biodiversity)and omits plantations (because of theirgeneral lack thereof).

The central question of the book—

Are things getting better?—is an impor-tant one. The reality is that significantprogress has been made in abating acidrain, although much still needs to bedone. And major efforts are under wayto stem deforestation and to address thetsunami of extinction. But it is crucial toremember that whereas deforestationand acid rain are theoretically reversible

(although there may be a threshold pastwhich remedy is impossible), extinctionis not. A dispassionate analysis, whichLomborg pretends to offer, of how farwe have come and how far we have yetto go would have been a great contribu-tion. Instead we see a pattern of denial.

The pattern is evident in the selectivequoting. In trying to show that it is im-possible to establish the extinction rate,he states: “Colinvaux admits in Scientif-ic American that the rate is ‘incalcula-ble,’” when Paul A. Colinvaux’s text,published in May 1989, is: “As humanbeings lay waste to massive tracts ofvegetation, an incalculable and unprece-dented number of species are rapidly be-coming extinct.” Why not show thatColinvaux thought the number is large?Biased language, such as “admits” inthis instance, permeates the book.

In addition to errors of bias, the textis rife with careless mistakes. Time andagain I sought to track references fromthe text to the footnotes to the bibliog-raphy to find but a mirage in the desert.

Far worse, Lomborg seems quite ig-norant of how environmental scienceproceeds: researchers identify a poten-tial problem, scientific examination teststhe various hypotheses, understandingof the problem often becomes morecomplex, researchers suggest remedialpolicies—and then the situation im-proves. By choosing to highlight the ini-tial step and skip to the outcome, he im-plies incorrectly that all environmental-ists do is exaggerate. The point is thatthings improve because of the efforts ofenvironmentalists to flag a particularproblem, investigate it and suggest poli-cies to remedy it. Sadly, the author seemsnot to reciprocate the respect biologistshave for statisticians.

Thomas Lovejoy is chief biodiversityadviser to the president of the WorldBank and senior adviser to the presi-dent of the United Nations Foundation.From 1973 to 1987 he directed theWorld Wildlife Fund–U.S., and from1987 to 1998 he served as assistant sec-retary for environmental and externalaffairs for the Smithsonian Institutionin Washington, D.C.

The Diversity of Life. E. O. Wilson. Belknap Press of Harvard University Press, 1992. (New edition. Penguin, 2001.)

Global Biodiversity Assessment. Edited by V. H.Heywood. Cambridge University Press, 1996.

Our Common Journey: A Transition towardSustainability. National Research CouncilBoard on Sustainable Development, PolicyDivision. National Academy Press, 1999.

Beyond Six Billion: Forecasting the World’sPopulation. Edited by John Bongaarts and Rodolfo A. Bulatao. National ResearchCouncil, 2000.

Climate of Uncertainty. George Musser inScientific American, Vol. 285, No. 4, pages14–15; October 2001.

Grand Challenges in Environmental Sciences,2001. National Research Council. Available atwww.nap.edu/catalog/9975.html

World Energy Assessment: Energy and theChallenge of Sustainability. United NationsDevelopment Program, United NationsDepartment of Economic and Social Affairs, andWorld Energy Council, 2001. Available atwww.undp.org/seed/eap/activities/wea/drafts-frame.html

The Intergovernmental Panel on ClimateChange Web site is available at www.ipcc.ch/

M O R E T O E X P L O R E

w w w . s c i a m . c o m
