7/27/2019 EJSD 2009

1/26

Electronic Journal of Sustainable Development

Te Electronic Journal of Sustainable Development(2009) 1(3)

Abstract

Tis paper examines whether over the long term, empiri-cal data supports Neo-Malthusian ears that exponential

population growth would lead to increasing resourcescarcity, and that increases in population, auence andtechnology would worsen human and environmental

well-being.It nds that, in act, global population is no longer

growing exponentially. Second, rom a historical per-spective, ood, energy and materials are more aordabletoday than they have been or much o human history.Tird, despite unprecedented growth in population,auence, consumption and technological change,human well-being has never been higher, and in the lastcentury it advanced whether trends in environmentalquality were up or down.

With respect to the environment, however, the recordis mixed. Initially, in the rich countries, auence and tech-nology worsened environmental quality, but eventuallythey provided the methods and means or cleaning up theenvironment. As a result, aer decades o deterioration,

their environment has improved substantially. Tat is,these countries have undergone an environmental tran-sition such that auence and technology are no longer

part o the problem, but are now part o the solution. Ingeneral, the world also seems to be on the verge o environ-mental transitions or cropland and water withdrawals.

Even developing countries have gone through theirenvironmental transitions or access to saer water andsanitation, and leaded gasoline. But these countries havenot yet made the transitions or other environmentalindicators in many places, although technological diu-

sion, accompanied by some auence, has moved themahead o where developed countries used to be at equiva-lent levels o development.

I the past is any guide, auence and technological

change are indispensable to ensuring that advances inhuman well-being continue into the uture even as envi-ronmental quality improves.

. Introduction

Concerns about population growth historically revolvedaround the notion that there may be insucient arableland, minerals or energy to meet the needs o an expo-nentially increasing population. Adding to these todayare ears that as technologies become more powerul and

wealth increases so too would consumption o naturalresources, which are urther compounded by worriesabout the wastes discharged to the air, land and water inthe course o developing and using these resources. Tus,the ear is that even i we do not run out o resources,

we might overwhelm the earths assimilative capacities.Absent empirical inormation, it can be plausibly argued

that together these actors conspire to increase environ-mental impacts with potentially disastrous eects onhuman welare.

Te general skepticism o population growth, eco-nomic development, and technology exhibited by many,i not most, environmentalists and Neo-Malthusians henceorth Neo-Malthusians is captured by the equa-tion, I = PA, where I is a measure o environmentalimpact, P is the population, A stands or auence asurrogate or per capita production or per capita con-sumption, oen measured in terms o the gross domestic

product (GDP) per capita and , denoting technology,is a measure o the impact per unit o production or con-sumption (e.g., Commoner ; Ehrlich and Holdren

Have increases in population, auence and technologyworsened human and environmental well-being?

Indur M. Goklany** Indur M. Goklany is the Assistant Director or Science and echnology Policy, Oce o Policy Analysis, US Department o the

Interior, and co-editor o theElectronic Journal o Sustainable Development.

Views expressed here are his, not those o any unit o the US government.

Email: editor =a= policynetwork.net (replace =a= with @)

7/27/2019 EJSD 2009

2/26

Have increases in popul ation, affluence and tecHnology worsened well-being?

; Ehrlich and Goulder ) . [echnology, as usedhere, includes both hardware (e.g., scrubbers, catalyticconvertors and carbon adsorption systems) and sowaretechnologies (e.g., policies and institutions that governor modulate human actions and behavior, culture, man-

agement techniques, computer programs to track ormodel environmental quality, and emissions trading)(Ausubel ; Goklany ).]

According to the IPA equation, i all else remainsthe same, an increase in population, auence or tech-nology would each act as multipliers or environmentalimpact (Commoner ; Ehrlich and Holdren ;Ehrlich and Ehrlich ; Ehrlich ). And as thatimpact increases, human well-being would necessarilydeteriorate.

Te IPA identity has been remarkably infuential.

It has intuitive appeal because o its apparent simplic-ity and seeming ability to explain how population, con-sumption or auence, and technology can aect humanand environmental well-being. It serves, or example, asthe master equation or the eld o industrial ecology(e.g., Graedel and Allenby ). One o its versionsunderpins the Intergovernmental Panel on ClimateChanges emission scenarios that have been used to esti-mate the amounts and rates o uture climate change andits impacts (IPCC , pp. ).

While noting that the IPA equation is a simpliedrepresentation and sometimes acknowledging that theterms on the right hand side are not independent o eachother, its ormulators have nevertheless used it to supporttheir contention that the human enterprise as currentlyconstituted is unsustainable in the long run, unless the

population shrinks, we diminish, i not reverse, over-consumption or economic development (particularlyin the United States), and apply the precautionary prin-ciple to new technologies, which in their view essentiallyembodies a presumption against urther technological

change unless the technology involved is proven sae andclean (Ehrlich and Holdren ; Ehrlich and Ehrlich; Myers ; Raensperger and ickner ).

Despite recognizing that benign technology couldreduce some impacts, many Neo-Malthusians argue, toquote Jared Diamond (, p.), it is a mistake tobelieve that [t]echnology will solve our problems. Inact, goes this argument, All o our current problemsare unintended negative consequences o our existingtechnology. Te rapid advances in technology duringthe th century have been creating dicult new prob-

lems aster than they have been solving old problemsDiamond (, pp. ). Ehrlich and co-workers arguethat or most important activities, new technology

would bring diminishing returns because as the bestresources are used up (e.g. minerals, ossil uels and armland), society would increasingly have to turn to mar-ginal or less desirable resources to satisy demand which

would increase energy use and pollution (Ehrlich and

Holdren ; Ehrlich et al. ).It has also been argued that technological advancescould be, and have been, counterproductive. First, suchadvances can reduce the cost o resource exploitation

which, then, increases environmental impacts orexample by using chain saws and bull dozers to createclear cuts o timber. Second, improved technologies canreduce the prices o consumption goods which stimulategreater demand and urther increase resource extraction.

Tere are alternative views o the impact o techno-logical change and economic growth regarding human

and environmental well-being. Simon, or instance,argues that economic growth and technological innova-tion conceived by more abundant brains in a more popu-lated world tend to improve human and environmental

well-being (Simon ). Others, such as Jesse Ausubel,argue that additional technological change has to be parto the solution, in order to reduce environmental impact,but nevertheless view economic growth as a multipliero impacts, rather than a contributor to the solution(Ausubel ; Landes ; Ryan ).

In the s, several economists undertook empiricalanalysis o the relationship between economic growthand environmental impact. Analysing a cross-sectiono countries, they ound that the relationship between

per capita GDP and environmental impact ollowed aninverted U shape similar to that identied by SimonKuznets or income inequality. On the basis o thisempirical relationship they posited the environmentalKuznets curve (EKC) hypothesis, which says that ascountries grow, the environment rst gets worse, then, asthey achieve a certain level o development, the damage

peaks and begins to improve again (Shak ; Gross-man and Krueger ).

In the environmental transition hypothesis, Goklany(, , a) has generalized the EKC hypothesisto attempt to account or both economic developmentand technological change. Under this hypothesis, ini-tially societies opt or economic and technological devel-opment over environmental quality because it allowsthem to escape rom poverty and improve their qualityo lie by making both needs and wants (e.g. ood, educa-tion, health, homes, comort, leisure and material goods)

more aordable. But once basic needs are met, over timemembers o society perceive that environmental dete-rioration compromises their quality o lie and they

7/27/2019 EJSD 2009

3/26

the electronic j ournal of sustainable development

www..g

start to address their environmental problems. Beingwealthier and having access to greater human capital,they are now better able to aord and employ cleanertechnologies. Consequently, environmental deteriora-tion can, rst, be halted and, then, reversed. Under thishypothesis, technological change and economic growth

may initially be the causes o environmental impacts, buteventually they work together to eect an environmen-tal transition aer which they become a necessary

part o the solution to environmental problems. Such atransition, i it occurs at all, would be evident as a peakin a stylized curve o environmental impact versus time,assuming that both economic development and technol-ogy advance with time. Tis assumption, while true ingeneral since Malthus time, hasnt always been so, noris there a guarantee that it will hold or all places at alltimes in the uture. Figure provides a stylized rendition

o the environmental transition hypothesis.In the ollowing, I examine whether long term empir-

ical data support the Neo-Malthusian notion that as

populations increase, become wealthier, and technol-ogy advances, we will run out o resources, leading toa deterioration o environmental quality, and human

well-being. I inspect trends that typically span severaldecades, because short term trends can be mislead-

ing. My examination, which is illustrative rather thanexhaustive, ocuses mainly on the U.S. because o thebetter availability and accessibility o long term data orthat country and because it has traveled urthest on the

path o economic development o any large economy. Inaddition, I use global data, where available, and also datarom a selection o less developed countries, mainly Indiaand China, in order to compare and contrast their expe-rience with that o the U.S.

With respect to human well-being, although I briefytouch on indicators such as poverty, education, child

labor, level o economic development, and economicand social reedom, I will use lie expectancy as the majorindicator o human well-being. Tis is consistent withits use as one o the three actors in the United Nationsoriginal Human Development Indicator (see e.g. UNDP). Some may object to the use o lie expectancyas an indicator o human well-being on the groundsthat a longer lie expectancy does not necessarily trans-late into better health. While theoretically this may becorrect, real world experience shows that as populationslive longer they also live more healthily, as evidencedby the act that the health-adjusted lie expectancy, i.e.,lie expectancy adjusted downward partially to discountlie years spent in poor health, is generally higher todaythan unadjusted lie expectancy in times past (Goklanya, p. ).

Regarding environmental quality, I examine trendsin various indicators o humanitys impacts on land, airand water. Specically, regarding the impact on land, Iuse cropland as the major environmental indicator, sinceconversion o habitat to cropland is generally deemed to

be the most signicant pressure on species, habitat andecosystems (see MEA ; Goklany , and reer-ences therein). With respect to water, I ocus on water

withdrawals and use because water diversion to meethuman needs is generally regarded as the greatest threatto reshwater biodiversity (e.g., IUCN , ;

Wilson ; see also MEA , p. ). Regarding airand water pollution, the selection o indicators is guidedlargely by the World Health Organizations ()analysis which estimates that water related diseases,indoor air pollution, and urban outdoor air pollution

are the largest environmental contributors to the globalmortality and disease burden. I also look at indicatorsrelated to climate change, not because it is among the

Figure 1 A stylized depiction of theEnvironmental Transition Hypothesis,a generalization of theEnvironmental Kuznets Curve

Race tothe bottom

region

C/Bregion

Environmentalstandard

p(T)

p(P) p(T)

Environmentalimpact(EI)

Time (or Technology or Affluence)

It shows the evolution of environmental quality the negative ofenvironmental impact (EI) as a society evolves from a low to ahigh level of economic development. The figure assumes thataffluence and technology advance with time, which is broadlyconsistent with historical experience since the start of the IndustrialRevolution. NOTE: p(P) = period of perception, the periodduring which the notion that environmental degradation cancompromise human well-being gains acceptance; p(T) = periodof transition, the period over which that perception leads toactions which eventually reduce environmental degradation;Race to the Bottom Region (where society strives to increaseeconomic development despite increasing EI); NIMBY Region =not in my backyard region (EI enters this region if benefits far

exceed costs to beneficiaries); C/B Region = cost/benefit region(where benefits and costs have to be more carefully balanced).Source: Goklany (2007a)

7/27/2019 EJSD 2009

4/26

7/27/2019 EJSD 2009

5/26

the electronic j ournal of sustainable development

www..g

Figure , based on cross country data rom the WorldBank (), shows that FR is inversely related to thelevel o economic development (as measured by GDP

per capita) and alls over time (a crude surrogate or tech-nological change)., Goklany (a, b) arguesthat the underlying relationships are more complex, withthe conditions supporting economic and technologicaldevelopment and, signicantly, the desire or such devel-opment, also important drivers.

First, since lower poverty the not-so-surprisingconsequence o economic growth means lower inantmortality rates and higher survival rates, it reduces pres-sures or more births. Tis is particularly importantbecause children are among the ew available orms oinsurance in poorer countries, which is one reason whythey have the highest FRs. Richer societies tend to have

social security programs which can reduce the pressureor more children. Second, higher incomes mean greateraccess to technology, which reduces the value o childlabor. Tird, richer societies oer greater educationaland economic opportunities or women, which alsoincreases the opportunity costs o their child bearing andchild rearing years. Fourth, the time and cost o educat-ing children to be competitive and productive in a richerand more technologically advanced society encouragessmall amily sizes.

Apart rom economic and technological develop-

ment, actors that contribute to economic growth andthe desire or greater wealth can help create conditionsthat tend to lower FR. In particular, literacy and the

amount o education, especially o women, helps propa-gate good habits o diet, nutrition, sanitation and saedrinking water. Tis improves health and reduces mor-tality, in general, and inant and maternal mortality, in

particular. As noted, this reduces pressures to maximize

birth and enables couples to plan the size o their ami-lies. At the same time, improved health leads to greaterwealth (or economic growth).

Finally, many couples arguably swayed by commer-cials and liestyles depicted by a ubiquitous, globalizedand globalizing visual mass media deer child birth inavor o current consumption (Goklany a).

ogether these actors explain why FR has droppedprogressively with both economic development andtime. Tus, in the IPA equation, P is not independento A and : sooner or later, as a nation grows richer, its

population growth rate alls (e.g., World Bank ),which might lead to a cleaner environment (Goklany, , b).

Tereore, while economic development and techno-logical change might initially increase the rate o popula-tion growth by reducing mortality rates, in the long run,they moderate population growth by helping directly orindirectly create the conditions or many amilies to vol-untarily opt or ewer children (and lower FR).

. Trends in human well-being

Although global population is no longer growing expo-nentially, it has quadrupled since . Concurrently,auence (or GDP per capita) has sextupled, global eco-nomic product (a measure o aggregate consumption)has increased -old and carbon dioxide has increasedover -old (Maddison ; GGDC ; WorldBank a; Marland et al. ). But contrary to Neo-Malthusian ears, average human well-being, measured

by any objective indicator, has never been higher.Food supplies, Malthus original concern, are up

worldwide. Global ood supplies per capita increasedrom , Cals/day in to , in (FAOSA). Tis helped reduce hunger and malnutri-tion worldwide. Te proportion o the population inthe developing world, suering rom chronic hungerdeclined rom percent to percent between and despite an percent populationincrease (Goklany a; FAO ).

Te reduction in hunger and malnutrition, along

with improvements in basic hygiene, improved access tosaer water and sanitation, broad adoption o vaccina-tions, antibiotics, pasteurization and other public health

Figure 2 Total fertility rate (TFR) vs. percapita income, 19772003

0.0

1.5

3.0

4.5

6.0

7.5

9.0

10.5

0 10,000 20,000 30,000 40,000 50,000 60,000

Childrenperwoman

GDP per capita (2000 international $, PPP-adjusted)

tfr1977p

tfr2003p

tfr1977

tfr2003

Source: Goklany (2007a)

7/27/2019 EJSD 2009

6/26

Have increases in popul ation, affluence and tecHnology worsened well-being?

measures, helped reduce mortality and increase lieexpectancies. Tese improvements rst became evidentin todays developed countries in the mid- to late-sand started to spread in earnest to developing countriesrom the s. Te inant mortality rate in developingcountries was per , live births in the early s;today it is . Consequently, global lie expectancy,

perhaps the single most important measure o humanwell-being, increased rom years in to years inthe early s to years today (Goklany a).

Globally, average annual per capita incomes tripledsince . Te proportion o the worlds populationoutside o high-income OECD countries living in abso-lute poverty (average consumption o less than $ perday in International dollars adjusted or purchasing

power parity), ell rom percent in to percent

in to percent in (Goklany a; WRI; World Bank ).

Equally important, the world is more literate and bettereducated. Child labor in low income countries declinedrom to percent between and . In mostcountries, people are reer politically, economically andsocially to pursue their goals as they see t. More peoplechoose their own rulers, and have reedom o expression.Tey are more likely to live under rule o law, and lesslikely to be arbitrarily deprived o lie, limb and property.Social and proessional mobility has never been greater. It

is easier to transcend the bonds o caste, place, gender, andother accidents o birth in the lottery o lie. People workewer hours, and have more money and better health to

enjoy their leisure time (Goklany a).Figure summarizes the U.S. experience over the th

century with respect to growth o population, auence,material, ossil uel energy and chemical consumption,and lie expectancy. It indicates that population hasmultiplied .-old; income, .-old; carbon dioxideemissions, .-old; material use, .-old; and organicchemical use, -old. Yet its lie expectancy increasedrom years to years and inant mortality (notshown) declined rom over per , live births to per ,.

It is also important to note that not only are peopleliving longer, they are healthier. Te disability rate orseniors declined percent between and /and, despite better diagnostic tools, major diseases (e.g.,cancer, and heart and respiratory diseases) occur

years later now than a century ago (Fogel ; Mantonet al. ).

I similar gures could be constructed or other coun-tries, most would indicate qualitatively similar trends,especially aer , except Sub-Saharan Arica and theerstwhile members o the Soviet Union. In the latter twocases, lie expectancy, which had increased ollowing

World War II, declined aer the late s to the earlys, possibly due poor economic perormance com-

pounded, especially in Sub-Saharan Arica, by AIDS,resurgence o malaria, and tuberculosis due mainly to

poor governance (breakdown o public health services)and other manmade causes (Goklany a, pp.,

pp., and reerences therein). However, there are

Figure 3 U.S. material, chemical and energy use, population and affluence compared to lifeexpectancy, 19002000

1

900

1

910

1

920

1

930

1

940

1

950

1

960

1

970

1

980

1

990

2

000

0

20

40

60

80

100

120

0

1

2

3

4

5

6

7

8

9

10

To

talma

teria

ls&organ

ics

(1900=

1)

lifeexpec

tancy

(years

)

Me

tals

,CO

2,

popu

lation,

affluence

(1900=

1)

Organics

Total materialsLife expectancy

Population (RH)

Metals (RH)

CO2

(RH )

A=GDP/cap (RH)

Source: Adapted from Goklany (2007a), based on Matos (2005), Marland et al. (2005), Maddison (2003)

7/27/2019 EJSD 2009

7/26

9

the electronic j ournal of sustainable development

www..g

signs o a turnaround, perhaps related to increased eco-nomic growth since the early s, although this could,o course, be a temporary blip (Goklany a; WorldBank a).

Notably, in most areas o the world, the health-adjusted lie expectancy (HALE), that is, lie expectancyadjusted downward or the severity and length o timespent by the average individual in a less-than-healthycondition, is greater now than the unadjusted lie expect-ancy was years ago. HALE or the China and Indiain , or instance, were . and . years, whichexceeded their unadjusted lie expectancy o . and. years in (WRI ).

Figure , based on cross country data, indicates thatcontrary to Neo-Malthusian ears, both lie expectancy

and inant mortality improve with the level o auence(economic development) and time, a surrogate or tech-nological change (Goklany a). Other indicators ohuman well-being that improve over time and as au-ence rises are: access to sae water and sanitation (seebelow), literacy, level o education, ood supplies percapita, and the prevalence o malnutrition (Goklanya, b).

4. Are ood and non-renewable resources

becoming scarcer?

Neo-Malthusians are also concerned that as populations

increase and become more auent, basic resources willbecome scarcer, and that we may even run out o some.Tis, o course, was the basis or the amous bet betweenPaul Ehrlich and Julian Simon over whether the priceo a basket o commodities would increase rom to

, which the latter won.However, in the last decade nominal (i.e., current)dollar prices or most commodities ood, energy, min-erals and metals have surged, due to increased demandand expansion o the money supply. In this section I willexamine whether and to what extent recent increaseshave made these commodities less aordable. I will ocuson metals, gasoline, and ood in a variety o settings.

My preerred index o aordability is the ratio o priceto an individuals wages or disposable personal income the lower the ratio, the less aordable the commodity.

However, where data on wages and disposable incomeare not available readily, I will use GDP or GNP percapita.

Metals

Figure shows trends rom to in the price othirteen metals or the U.S. relative to wages, indexed tothe level (=). Tis indicates that these metalsare generally priced higher today than in the s, butnot as high as they were in the s and s (except

possibly or zinc and nickel). Perhaps more importantly,they are more aordable today than or most o history.O course, we have no idea whether the current blips willbecome a long term trend or recede like previous blips inthe long slide in prices-relative-to-wages.

Figure shows indices or the nominal and real priceo metals rom to . Te nominal price indexis patched together rom Paenzellers () indexor six metals (aluminum, copper, lead, silver, tin, and

zinc) or , and World Banks (c, )metals and minerals price index or the remainder o the

period. Te real price index is derived rom the nominalprice index using the BEA () GDP defator or, and the implicit price index published inU.S. Bureau o the Census (: ) rom .Although the indices are patched together using dier-ent data sources, there is no escaping the surge in pricessince . Even in real terms, the metals priceindex hasnt been higher since World War I. However,in terms o aordability, estimated as the real GDP per

capita divided by the real price, the picture is a littledierent.

Figure shows that aordability peaked in . It

Figure 4 Life expectancy & infant mortality asa function of economic developmentand secular technological change,19601999)

Life expectancy, 1999

Life expectancy, 1960

Infant mortality, 1999

Infant mortality, 1960

Note: MXR = market exchange rates.Source: Goklany (2007a), based on World Bank (2001)

0 10,000

20,000

30,000

40,000

0

50

100

150

200

250

300

350

400

450

0

10

20

30

40

50

60

70

80

90

Infantmortality(deaths/1,0

00births)

Lifeexpectancy(years)

7/27/2019 EJSD 2009

8/26

10

Have increases in popul ation, affluence and tecHnology worsened well-being?

is presently at the level or the United States,and the level or India. So, despite recent

price run ups due to unprecedented demand, metals aremore aordable today than they have been or much ohistory. For the average Indian, metals are eight timesmore aordable today than in , and or the averageAmerican, they are thirteen times more aordable.

Food

Essentially similar patterns as that or metals are evidentor ood aordability (see Figure , which has beendeveloped using the same sources and methods as thatused or the previous gure). Food aordability peakedin or both the U.S. and India; ood is timesmore aordable today or the average American than it

Figure 5 Metal prices relative to wages, U.S., 18002007

1800

1810

1820

1830

1840

1850

1860

1870

1880

1890

1900

1910

1920

1930

1940

1950

1960

1970

1980

1990

2000

2001

2002

2003

2004

2005

2006

2007

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

Pricerelativetowages(indexedto199

0=1

00) Aluminium

Antimony

Copper

Lead

Magnesium

Manganese

Mercury

Nickel

Silver

Platinum

Tin

Tungsten

Zinc

Sources: (1) Data for 18001990 are from Moore (1995). (2) Price data from 19902007 are from various issues of USGS, MineralCommodities Summaries and Minerals Year Books, available at http://minerals.usgs.gov/minerals/pubs/commodity/, visited on July 7, 2008.(3) Wage data for 19902007 are from Bureau of Labor Statistics, Establishment Data: Historical Hours and Earnings, available at,ftp://ftp.bls.gov/pub/suppl/empsit.ceseeb2.txt, visited June 27, 2008

Figure 6 Metals commodity indices, 19002008

1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000

0

100

200

300

400

500

600 Nominal US $

Constant 2000 $

Sources: (1) Commodity price in nominal dollars is indexed to 100 in 2000. Data are from Pfaffenzellers (2007) index for metals for

19002000, and the World Bank index for minerals and metals for 20012008, courtesy of Betty Dow (World Bank 2008c) and World Bank(2009). (2) Constant dollars are based on GDP deflator for the U.S. from 19292008 using BEA (2009), and GNP deflator from USBC(1975:224), for 19001928

MetalsIndex(2000level=

100)

7/27/2019 EJSD 2009

9/26

11

the electronic j ournal of sustainable development

www..g

was in , whereas or the average Indian it is timesmore aordable. Tis is one actor in the increased avail-ability o ood supplies per capita in India, and the longterm decline in the proportion o the Indian populationsuering rom hunger and malnutrition (see Goklanya, p. ).

Gasoline in the U.S.

Figure shows rom to mid-, the nominaland real price indices or gasoline, and a gasoline aord-ability index or the U.S., the last calculated as the ratioo the average persons disposable income to the priceo gasoline. Tis gure, which uses a nominal price oregular gasoline o $. a gallon in , shows thatboth the real infation-adjusted price and the nominal

Figure 7 Metals affordability index, India and U.S., 19002008

1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000

0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000 Metal affordability, USA

Metal affordability, India

Sources: (1) See Figure 5. (2) U.S. and Indian GDP per capita are based on Maddison (2003) for 19001980, and GGDC (2008) for19812007. For 2008, U.S. GDP per capita is based on the 20072008 growth in GDP per capita from BEA (2009), while Indian GDP percapita is based on a 5 percent growth in GDP per capita from 2007 to 2008 per EIU (2008)

FoodaffordabilityindexedtoUS=1

00

in1900

Figure 8 Food affordability index, India and U.S., 1900June 2008

1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000

0

500

1,000

1,500

2,000

2,500 Food affordability, USA

Food affordability, India

Sources: (1) Commodity price in nominal dollars is indexed to 100 in 2000. Data are from Pfaffenzellers (2007) index for metals for19002000, and the World Bank index for minerals and metals for 20012008, courtesy of Betty Dow (World Bank 2008c) and World Bank(2009). (2) Constant dollars are based on GDP deflator for the U.S. from 19292008 using BEA (2009), and GNP deflator from USBC(1975:224), for 19001928

7/27/2019 EJSD 2009

10/26

12

Have increases in popul ation, affluence and tecHnology worsened well-being?

price in were at the highest they had been since atleast . Te gasoline aordability index (indexed tothe level = ) peaked in at .. During ,averaged over the year, it was at ., a level rst reachedin and last seen in .

5. Trends in environmental well-being

In this section I will examine long term trends in variouskey environmental indicators to establish whether theyare consistent with Neo-Malthusian or other viewsregarding the eect o economic growth and techno-logical change on the environment. I will use the IPAequation to determine how well changes in impacts (I)

track with changes in population (P), auence (A), andtechnology ().

Estimating technological change

In applying the IPA equation, auence will be meas-ured by GDP per capita or, i thats unavailable, grossnational product (GNP) per capita. For the U.S. the di-erence between these two measures in any year is slight on average, within . percent (with a range rom

+. to -. percent) or (Goklany a).Since A is represented by GDP per capita, the IPA

equation may be rewritten as:

I population (GDP/population) ()

Since total consumption the product o P and A is equivalent to GDP, the technology-actor () can beestimated using:

= I/GDP ()

Tus, is equivalent to impact per unit o GDP.Notably, a decline in would reduce I and denotes animprovement in technology.

Te technological change () rom an initial time(ti) to nal time (t) can then be estimated by:

= (I/GDP) ()

I population, auence, their product (GDP), andthe technology-actor are all normalized to unity at ti,then

= (I/GDP ) , ()

where subscript denotes the value at the end o theperiod.

Where emissions (E) are used to characterize the envi-ronmental impact, technological change is the change inemissions per GDP, that is,

=(emissions/GDP)= (E/GDP) ()

I will, except where noted, use Equation (or )to estimate technological change, and whether thathas made matters better or, consistent with the Neo-Malthusian view, worse over the period o analysis. Forsome indicators, e.g., mortality rom extreme weatherevents (a purported indicator o global warming) or

water related diseases, I will substitute P or GDP

(= P x A) in the above equations on the basis that,ceteris paribus, as a rst order approximation, mortalityincreases linearly with P but is relatively insensitive toauence.

5.1 Cropland or terrestrial habitat conversion

Because cropland is critical or producing the ood andnutrition necessary to ward o hunger and malnutrition still among the largest contributors to global mortal-

ity (Goklany a, pp. ) the rst Malthusianconcern was that humanity may run out o cropland.Now many are concerned that there may be too much

Figure 9 Gasoline affordability index,19492008, (1960=1)

1949

1957

1965

1973

1981

1989

1997

2005

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0Nominal price

Real price (in 2000 US $)

Affordability Index (RH axis)

The affordability index is calculated as the ratio of per capitadisposable income to the price per U.S. gallon of regular gasoline(see text).Sources: DOE (2009), BEA (2009)

Nominal&realprice($/gallon)

Affordabilityindex(1960=1

)realdispo

sable

income-to-realprice

7/27/2019 EJSD 2009

11/26

13

the electronic j ournal of sustainable development

www..g

cropland. In act, the single largest threat to terrestrialecosystems and biodiversity is the diversion o habitat toagricultural uses, particularly cropland (Goklany ;MEA : ).

Figure shows trends or the U.S. rom to

in the amount o cropland planted, population (P),auence (A), GDP (= P x A), as well as two measureso technology, namely, (calculated as I/GDP) and (calculated as I/P). All variables are normalized to .

Tis gure shows that despite a more-than-triplingo the population and a -old increase in consumption(GDP), cropland was unchanged at million acres.Tat is, the impact as measured by this indicator has notincreased, contrary to nave interpretations o the IPAequation. Tis is because the decline in the -actor hascompensated or these increases. was at . in

relative to , i.e., technology reduced the impact by percent over what it would otherwise have been.

Arguably, however, it is more appropriate to useGDP (= P x A) to estimate technology because auenceincreases the demand or meat and milk, and the pro-

pensity or wastage. Using this measure, (= I/GDP)stands at . in relative to , that is technol-ogy reduced impact by percent. Perhaps, the correctmeasure would be to use the product o population andthe logarithm o auence. Regardless, and bracketthe range or technology.

Note that cropland was higher ( million acres) inboth and than in and . Te currentarea o cropland might have been lower still but or subsi-dies which have partially negated the improvements thattechnological change might otherwise have achieved.Note, however, that some o the increase in yield thathas helped halt land conversion could be due to highercarbon dioxide concentrations (e.g., IPCC , p.).Tis is included in the technology term by deault.

Figure , which shows global trends in cropland rom

to , also oers no support or the propositionthat increases in population and auence necessarilyincrease impacts. In act, this gure indicates that tech-nological change since saved about , millionhectares rom conversion to cropland and that, like inthe U.S., cropland may be peaking globally (that is, goingthrough an environmental transition; Goklany a).

Whether it actually stabilizes and/or declines consistentwith the environmental transition hypothesis dependson the availability, and barriers to, technological change.In this regard, European attitudes toward genetically

modied crops, and the diusion o those attitudesto developing countries, particularly in Arica, retardstechnological change and are counterproductive, as are

subsidies in developed countries which keep more landunder cultivation (Goklany a).

Figure 10 IPAT for U.S. cropland, 19102006

1910

1920

1930

1940

1950

1960

1970

1980

1990

2000

0

4

8

12

16

20

24

0.0

0.5

1.0

1.5

P,

A,

PxA(1910=1

)

Impact&technology,

1910=

1

PAPxA

)

T (RH)T' (RH)I (RH)

Sources: GDP and population data are from Maddison (2003),GGDC (2008), World Bank (2008a); cropland data is from ERS(2008)

Figure 11 Global cropland (in billion hectares),population (in billion people),affluence (in thousands of 1990,PPP-adjusted International$), andcropland per capita, 17002005

0

1,000

2,000

1700 1800 1900 2000

3,000

4,000

5,000

6,000

7,000

0.0

0.1

0.2

0.3

0.4

0.5

Cropland,population,

affluence

Cropland

percapita

Cropland

Population

AffluenceCropland per capita (RH)

Cropland, no tech change

Note that the difference between the dotted and solid blue linesequals the amount of habitat saved from conversion hadtechnology been frozen at 1961 levels.Sources: Goklany (2007a), Maddison (2003), GGDC (2008),FAOSTAT (2008); World Bank (2008a)

7/27/2019 EJSD 2009

12/26

1

Have increases in popul ation, affluence and tecHnology worsened well-being?

5.2 Water withdrawal and consumption

Just as the diversion o land to meet human needs isthe single greatest threat to terrestrial biodiversity, sois diversion o water the greatest threat to reshwater

biodiversity.For the United States, over the -year period,, the split between surace and ground water

withdrawals has stayed more or less constant at /percent, respectively (Hutton et al. ). Te portiono surace-water withdrawals that was classied as salineincreased rom to percent between and . Ithas since remained approximately constant.

Between and , while U.S. populationincreased by percent and, economic consumption by percent, total water withdrawals increased by

percent. However, between and , water with-drawals declined by percent despite increases o and percent in population and economic consumption,respectively (Hutson et al. ). Moreover, the longterm trend o declining total wetland area in the U.S.seems to have halted and even reversed, with ,acres being added between and (Dahl ).

By contrast with the reduction in water withdrawalsin the U.S., data rom Shiklomanov () indicates that

while water withdrawals and use might be approach-ing saturation globally, they had not peaked as o ,although on a per capita basis, they began to decline inthe s. See Figure .

5.3 Water-related impacts

Water has traditionally been high on the list o environ-mental priorities because o the potential o death anddisease rom water related diseases. Figure shows thatrom , U.S. death rates due to various water-

related diseases dysentery, typhoid, paratyphoid, othergastrointestinal disease, and malaria declined by . to. percent (USBC, various years).

Tese reductions, which preceded the CleanWater Act, can be attributed to, among other things,greater knowledge o better hygiene, greater access tosae water and sanitation, and new and more eectivetherapies.

Analysis o cross country data indicates that with eco-nomic development and time, access to both sae waterand sanitation generally increases in terms o absolute

numbers and, more signicantly, as a proportion o totalpopulation (Goklany a, b). Because o higherlevels o economic development and technological

diusion, such access, although not yet universal, hasnever been higher. Between and the early s,or example, the proportion o the population withaccess to saer water increased rom to percent inSouth Asia and to percent in Sub-Saharan Arica,

while with regard to sanitation, it increased rom to percent in South Asia, and to percent in Sub-Saharan Arica (World Bank b).

Figure 12 Global water withdrawals and use,

19001995

Withdrawal

Use

PopulationWithdrawal/cap

Use/cap

1900

1920

1940

1950

1960

1970

1980

1990

2000

0

1

2

3

4

5

6

7

Source: Shiklomanov (2000)

Figure 13 Death rates for various water

related diseases, 19001970

0

200

400

1900 1910 1920 1930 1940 1950 1960 1970

600

800

1,000

1,200

1,400

1,600

0

50

100

150

200

250

300

350

400

Gastrointestinaldiseases

Dysentery,malaria,

typhoid&paratypho

id

Dysentery

Malaria

Typhoid and p-typhoid

Gastrointestinal disease

Source: Goklany (2007a), based on USBC (various years, 1975)

7/27/2019 EJSD 2009

13/26

1

the electronic j ournal of sustainable development

www..g

5.4 Traditional air pollution

Concern over traditional air pollutants soot, otherorms o particulate matter, sulur dioxide, carbon mon-oxide and, in some places, ozone was instrumental

in raising environmental consciousness in the U.S. andtodays richer countries. Long term data indicates thatair quality or these traditional pollutants has generallyimproved, particularly or the substances and in theareas that were o the greatest public health concern(Goklany a). For these countries, long term airquality trends show pronounced peaks that are generallyconsistent with the environmental transition hypothesisrather than with Neo-Malthusian theories that auenceand technological change make matters worse.

With respect to the U.S., probably a harbinger or

other countries, the earliest environmental transitionsapparently occurred or indoor air quality (by the s),ollowed later by improvements in outdoor air quality.Tis is especially signicant because the vast major-ity o people spend the majority o their time indoors,generally at home. Tereore indoor exposure is perhapsthe single most critical determinant o the potential

public health impact o air pollution. Remarkably, theseimprovements in indoor air quality, which were enabledby improvements in technology and greater auence,occurred voluntarily as households moved away romsolid uels such as coal and wood to cleaner energysources within the home oil, gas, electricity.

With respect to U.S. national outdoor air quality aswell, the transitions seem to have occurred earlier orpollutants and locations that were o the earliest andgreatest concern. Tey occurred rst or total suspended

particulate matter (around ), ollowed by sulurdioxide (early-to-mid s), carbon monoxide (mid-to-late s), lead (mid-to-late s), ozone (mid-to-lates nationally but mid-s in Caliornia, where

it was a major early concern), and nally nitrogen oxides(in the late-s). Perhaps surprisingly, many o thesetransitions also preceded the U.S. Clean Air Act Amend-ments o .

For the traditional air pollutants, trends in emis-sions (an indicator o less relevance to public health and

welare than either indoor or outdoor air quality), indi-cate that they too have gone through their environmen-tal transitions in the US.

Notably, air quality in the currently industrializing(or developing) countries is substantially worse than in

developed countries. Beijing, Mexico City, New Delhiand Cairo, or instance, are among the most pollutedcities in the world. Nevertheless, developing countries

seem to have learnt rom the experience o developedcountries. In act, in many respects, they are ahead o

where industrialized countries were when they were atthe same level o economic development. For example,the U.S. rst introduced unleaded gasoline in

when its GDP per capita was $,, whereas Indiaand China instituted some controls or lead-in-gasolineby , when their GDPs per capita were $, and$,, respectively (Maddison ; Goklany a).By , only about countries out o about wereusing leaded gasoline (Dumitrescu ) although theglobal GDP per capita was about $,. Tis, o course,is due to the diusion o knowledge and technologyrom industrialized to developing countries.

Analysis o air pollution trends rom in twenty Asian cities Bangkok, Beijing, Busan,

Colombo, Dhaka, Delhi, Hanoi, Ho Chi Minh, HongKong, Jakarta, Kathmandu, Kolkata, Mumbai, Manila,Seoul, Shanghai, Singapore, Surabaya, aipei and okyo showed that, in general, SP and PM- decreasedbetween and , although ambient levels wereabove limits set by the World Health Organization(Figure ; CAI-Asia ). [Sixteen o the cities arein developing Asia.] For SO, levels had been improvedto within WHO guidelines. For NO, the levels hadbeen stabilized around WHO guidelines. Tese resultsare consistent with Hao and Wangs () analysisindicating that despite substantial emission increases,average concentrations in Chinese cities or total sus-

pended particulates, PM-, and SO declined by , and percent, respectively, between and .

Figure 14 Air quality trends in 20 major Asiancities, 19932004

Source: CAl-Asia (2006)

0

50

100

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

150

200

250

300 TSPPM10

NO2

SO2

7/27/2019 EJSD 2009

14/26

1

Have increases in popul ation, affluence and tecHnology worsened well-being?

In other words, some areas in Asia have apparently gonethrough their environmental transitions or a variety oair pollutants, and at lower levels o economic develop-

ment than in the U.S.Tere have been improvements in Latin America as

well. Figure shows improvements in Mexico City, leg-endary or its air pollution, rom (Molinaet al. ). Similarly, PM- concentrations in Brazilsindustrial region o Cubatao among the worlds astestgrowing industrial areas declined rom to about micrograms per cubic meter rom (Wheeler). Fine particulate matter (PM-.) concentrationsdropped percent in Santiago, Chile between and (Koutrakis et al. ).

Te major air pollution problems in developing coun-tries are, however, indoors. Hal o the worlds popula-tion continues to use solid uels such as coal, dung and

wood. Te World Health Organizations Global Burdeno Disease 2000 (Version 2) study estimates that in air pollution was responsible or . million premature

deaths (or . percent o all deaths). wo-thirds o thesedeaths were attributed to indoor pollution rom particu-late matter in developing countries rom cooking andheating with coal, dung and wood and the remainderto outdoor air pollution (WHO a, b; Bruceet al. ). On the basis o disability-adjusted lie years(DALYs), a measure which discounts every year o lielived under a disability by the severity o that disability,indoor air pollution accounts or . percent o annuallost DALYs worldwide, and outdoor air or . percent.

I the currently poor inhabitants o less developed

countries were to grow richer, they would have themeans to switch out o dirty solid uels and into cleanerestablished technologies such as natural gas, oil or even

Figure 15 Air quality trends in the Mexico City metropolitan area, 19862006

Source: Molina et al. (2008)

1986

1991

1996

2001

2006

0

2

4

6

8

10

0.0

0.1

0.2

0.3

0.4

CO(

ppm)

NO

x(ppm)

NOx

NO2

TSP

PM10

PM2.5

SO2

CO

1986

1991

1996

2001

2006

0.00

0.05

0.10

0.15

0.20

0.25

0.30

O2

(ppm)

1986

1991

1996

2001

2006

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.00

0.04

0.08

0.12

0.16

0.20

SO

2

(ppm

)

NO

2

(ppm

)

1986

1991

1996

2001

2006

0

200

400

600

800

1,000

0

50

100

150

200

250

300

TSP(ug/m3)

PM10,

PM2.5

(ug/m3)

7/27/2019 EJSD 2009

15/26

1

the electronic j ournal of sustainable development

www..g

electricity, which would help reduce the disease burdenin these countries signicantly. It would essentially allowtodays developing countries to ollow the same path sosuccessully taken by the rich nations in reducing popu-lation exposure to air pollutants. Tis is essentially the

opposite o the claim made by opponents o auence.

5.5 Global warming

Carbon dioxide

Unlike the other environmental indicators examinedthus ar, carbon dioxide has only recently been elevatedin the popular mind as a signicant environmental

problem. Arguably, this elevation did not occur until

around the late s with the passage o the Kyoto Pro-tocol or even the rst decade o the st century, withthe publication o the IPCCs Tird and Fourth Assess-ment Reports. Even now, some would dispute this char-acterization, while others would dispute its importance(e.g., Lomborg ; Goklany ). Eorts to reduceCO, thereore, are still immature. Tis task is urthercomplicated by the socioeconomic consequences oreducing CO emissions (Nordhaus ) and the actthat it will necessarily take time, technology and capitalto modiy existing energy inrastructure. Accordingly, itis no surprise that empirical trends do not indicate thatCO has peaked. However, in many places emissions

per GDP (or the carbon intensity) have peaked and arealling steeply. Indeed, C per unit o GDP is a leadingenvironmental indicator (because absent a long term sus-tained reduction in it, a growing economy will be unableto bring about a transition with respect to total emis-sions; Goklany a).

Figures and show U.S. and global trends rom to in each o the terms o the IPA equation

and PxA (=GDP), all normalized to or C. I(i.e., CO emissions), P, A, and PxA are plotted on thele hand axis, and technology (=I/PxA) on the righthand axis.

Tey show that or the U.S., despite a -old increasein GDP since , CO emissions increased -old.Tis translates into a percent reduction in impact

per unit o consumption (i.e, the -actor, which isalso the carbon intensity o the economy) during this

period, or a . percent reduction per year in the carbonintensity between and . Since , however,

U.S. carbon intensity has declined at an annual rate o. percent. (Arguably, CO emissions might have beenlower, but or the hurdles aced by nuclear power.)

Globally, output increased -old since , whileCO increased -old because technology reduced theimpact cumulatively by percent or . percent per

year. Both U.S. and global carbon intensity increaseduntil the early decades o the th century. Since ,global carbon intensity has declined at the rate o .

percent per year.Figure shows the components o IPA and PxA or

China rom , each normalized to levels.Chinese output increased by a actor o since ,

Figure 16 U.S. IPAT trends for CO2, 18202004

I

P

A

PxA

T (RH)

Sources: Marland et al. (2007), Maddison (2003), GGDC (2008)

1810 1860 1910 1960 2010

0

5

10

15

20

25

30

0.0

0.5

1.0

1.5

I,P,

A,

PxA(

1900=

1)

Technology(=c

arbonintensity)

(emissionsperGDP,

1900=1

)

Figure 17 Global IPAT trends for CO2,

18202004

0

5

10

15

20

1810 1860 1910 1960 2010

25

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

I,P

,A

,Px

A(1900=

1)

Tec

hno

logy

(=c

arbon

intensity

)

(em

iss

ionsper

GDP

,1900=

1)

Total CO2

emissions fromfossil-fuels (million metric tons of C)

Population

AffluencePxA

T-factor

Sources: Marland et al. (2007), Maddison (2003), GGDC (2008),World Bank (2008a)

7/27/2019 EJSD 2009

16/26

1

Have increases in popul ation, affluence and tecHnology worsened well-being?

while CO emissions increased by more than twice asmuch (a actor o ), refecting its rapid transition roma rural agrarian society to the worlds manuactory. From to , improvements in carbon intensity ailed

to keep pace with the cumulative output increase. Since, the rst year o Chinas economic reorms, carbonintensity has dropped at an annual rate o . percent,in part because o the reorms. Nevertheless, this drophas not been large enough to compensate or the tremen-

dous increase in consumption.Since improvements in technology mostly precededgeneral recognition o CO as an environmental issue,they can mainly be attributed to business as usual

where all economic participants seek to maximizeprivate welare (including prots) via minimization ocosts (including energy costs).

Deaths due to extreme eents

o the extent that extreme weather events are exacer-bated by global warming, deaths due to such eventscould be an indicator o the impacts o global warming.

Globally, both cumulative mortality and cumulativemortality rate or all extreme weather events (namely,drought, extreme temperatures, foods, slides, waves andsurges, wildres, and wind storms) have been decliningsince the s [Goklany (c), based on data rom

Figure 18 China IPAT trends for C02,

19002004

0

10

20

30

1900 1915 1930 1945 1960 1975 1990 2005

40

50

60

70

0

1

2

3

4

5

6

I,P,

A,

PxA(1950=1

)

Technology(=

carbonintensity)

(emissionsperGDP,

1950=1

)

I

P

A

PxA

T (RH)

Sources: Marland et al. (2007), Maddison (2003), GGDC (2008)

Table 1: The eects o technological change on declining U.S. and global mortality and mortality

rate or extreme weather events during the 20th century

Deaths,

earliest 10-

year period

Deaths, nal

10-year

period

Death rates,

earliest 10-

year period

(per million)

Death rates,

nal 10-year

period

(per million)

Technological

change

(%), based on

mortality rate

Rate o

technological

change (%/

year), based

on mortality

rate

World

all extreme events

1900/091997/2006

1,280,000 258,000 78.8 4.17 95.3 3.1

U.S.Hurricanes

1900/091997/2006

8,730 1,760 11.3 0.60 94.7 4.6

U.S.

foods

1903/121997/2006

260 740 0.31 0.26 15.8 0.3

U.S.

tornados

1917/261997/2006

3,160 620 2.90 0.22 92.5 4.1

U.S.

Lightning

1959/681997/2006

1,180 440 0.63 0.16 75.4 2.2

Source: Goklany (2007c), based on EM-DAT (2007) or global mortality data; McEvedy and Jones (1978) or global population; Blake etal. (2007) or U.S. hurricanes; NCDC (2005, 2007) and NWS (2007) or U.S. lightning and tornados; HIC (2007) or U.S. foods; USBC

(2007 or U.S. population. NOTE: A negative sign indicates that technological change reduces impacts.

7/27/2019 EJSD 2009

17/26

19

the electronic j ournal of sustainable development

www..g

EM-DA ()]). While older data are necessarilysuspect, between to , mortalityapparently dropped by percent and mortality rate by percent, the latter at an annual rate o . percent (seeable ). Te drops aer the s (not shown below) are

even steeper (Goklany c).able also shows that with regard to the U.S., asimilar comparison o the earliest -year period to thelatest -year period () or which data wereavailable at the time o analysis, mortality due to hur-ricanes, tornados and lightning was reduced by , and percent, and mortalityrate by an annual rate o., . and . percent per year, respectively. However,or foods, mortality increased by percent, but mor-tality rate declined percent (an annual rate o .

percent).

Note that or each o these U.S. indicators in able, except hurricanes, mortality and mortality rates

peaked during the th century. For hurricanes the peakoccurred in , and was dominated by the ,atalities due to the Galveston hurricane, and there

was a subsidiary peak during the last period because othe hurricane season, or which I used a death tallyo , per Blake ().

6. Efects o long term technological change onimpacts

able shows or the environmental indicators and areasexamined in Section , long term changes in environ-mental impact (I), population (P), auence (A), their

product (GDP = P x A), the technology actor (), andtechnological change (). and are calculatedusing Equations or , except or mortality, where popu-lation is substituted or GDP.

Te entries or each o the components o the IPA

equation are their values at the end o the period o anal-ysis normalized to unity at the beginning o the period.Tus, the rst row indicates that in , U.S. population

was .-times its level; auence, .-times; GDP,.-times. Nevertheless, the environmental impact oU.S. agriculture, measured by the amount o cropland,

was essentially unchanged. , measured by cropland perGDP, was . times its level. Hence, the amounto technological change () during the intervening

period the percent change in impact per unit o GDP is minus . percent (in the second last column). [Te

minus sign indicates that the environmental impact perunit o GDP declined, i.e., matters improved.] Finally,the last column provides an estimate o the annual rate

o technological change, assuming exponential change(minus . percent per year).

As with all trends, results displayed in able canbe sensitive to the starting and ending years used orcompiling the data, particularly or episodic events, e.g.,

extreme weather events. o avoid bias, in these cases Iused the longest readily available record.Tis table indicates that since auence has

increased aster than population worldwide, and in theU.S., China and India.

Second, but or technological change, impacts wouldgenerally have been much higher, in many instances byan order o magnitude or more. For instance, per unit oGDP, technological change reduced the global environ-mental impact o agriculture by percent rom to . In act, it has stabilized the amount o habitat

converted to cropland in the U.S. and almost stabilized itglobally (Figures and ). During the th century, itreduced death rates rom various water related diseases inthe U.S. by . percent. It also reduced the cumu-lative global death rate rom extreme weather events by percent, while reducing U.S. death rates rom hurri-canes, lightning, foods and tornados by percent.Because o technology, U.S. indoor air pollution levelsare currently to (+) percent lower than they oth-erwise would be. However, while technology reducedthe rate o increase, CO emissions, nevertheless, grewsubstantially.

Tird, improvements are apparently more pro-nounced or indicators most directly related to human

well-being. Specically, or each pollutant, indoor airquality improved earlier and aster than outdoor emis-sions (which comprise the bulk o emissions), and mor-tality rates were reduced more than indicators whoserelationship to public health is more indirect. Withrespect to global warming related indicators, mortalityrates rom total extreme weather events declined sub-

stantially, although carbon dioxide emissions increaseddespite reductions in the carbon intensities o econo-mies. Te latter is true even in India and China, whererecent improvements in carbon intensities coincide withthe initiation o economic liberalization, despite gener-ous uel subsidies to consumers.

For the environmental indicators used to characterizethe impacts on land, air, and water cropland, indoor airquality, traditional air pollutant emissions, and mortalityrom water-related diseases technological change gen-erally more than compensated or any long term increase

that might have occurred in impact due to increases ineither population or auence, but not always or thecombined eect o the two (i.e., P x A). Te exceptions

7/27/2019 EJSD 2009

18/26

20

Have increases in popul ation, affluence and tecHnology worsened well-being?

to this are: (a) U.S. NOx emissions where technologycompensated or population increase between and, but not or auence, (b) water withdrawals or theU.S. rom , where technology compensatedor population but not or auence, and (c) global water

withdrawals and consumption rom , wheretechnology ailed to keep pace with either populationor auence.

What the table does not show is that even where tech-nology was unable ully to compensate or the increasein aggregate output over the entire period water with-drawals and national air emissions are cases in point itmoderated impacts so that, by the end o the period, inmost cases impacts had peaked and were substantiallylower than in previous decades (Goklany a, p. ).

In general, long term environmental trends have not

conormed to the notion that, sooner or later, technol-ogy will necessarily increase environmental impacts.Moreover, i one goes suciently ar back into the his-torical record, e.g., or habitat converted to cropland, air

pollution emissions or water related diseases, the initialtrends will show environmental deterioration, seemingly

validating the Neo-Malthusian view. But over time thisinterpretation ails, as the environmental impact is moreor less halted (e.g., cropland) or even reversed (air and

water pollution) (Goklany a). Such declines lendcredence to the environmental transition hypothesisand indicate that, in eect, sooner or later technologyno longer acts as a multiplier, but as a divisor or the envi-ronmental impact.

7. Discussion

Long term empirical trends oer little support or Neo-Malthusian worldviews. Yes, global population has con-tinued to rise, as has auence, output and consumption,

But metals, ood and energy are more aordable todaythan they have been or much o history. More impor-tantly, human well-being has never been higher. Moreo-

ver, population is no longer increasing exponentially. Inact, there are signs that it could plateau, and possiblyeven decline in the coming decades.

Initially in the arc o development, environmentalquality indeed suered, but by virtually every criticalmeasure hunger, malnutrition, mortality, education,income, liberty, leisure, material goods, mobility, lieexpectancy human well-being advanced. In the U.S.,

or instance, this advance has been more or less continu-ous since the s, despite the waxing and waning o a

variety o environmental problems in the interim. And

this is also true or the world as a whole, at least since thes (Goklany a).

Te improvements in human well-being despiteincreased population suggest that contrary to Neo-Malthusian claims e.g., Diamond (, p. ), au-

ence and technology have solved more problems thanthey have created.Historically, in the richer countries hunger and water

related diseases were conquered rst, then indoor airpollution, and nally outdoor air pollution. Once richercountries learned to cope with water related diseasessuch as cholera and dysentery (through knowledge obasic hygiene, a better understanding o the causes othese diseases, better access to sae water and sanita-tion, draining o swamps, and so orth), there was little

public emphasis on other environmental problems.

Despite that, private actions or the most part cleanedup indoor air pollution. Tese actions, including volun-tary switching to cleaner uels and installation o moreecient combustion appliances, were enabled by greater

prosperity and technological change, and driven by eachhouseholds natural desire to advance its own quality olie (Goklany a, pp. ).

Similar economic and behavioral orces were also atwork or outdoor air pollution, and the pollution inten-sity o the economy declined, but not rapidly enough eventhough, in retrospect, many o the traditional air pollut-ants were in the midst o, or had even gone through, theirenvironmental transitions (Goklany a, pp. ,, ). In the U.S., in the wake o the prosper-ity o the s and early s and once the privationso the Great Depression and World War II had becomedistant, the clamor or governmental intervention grew.Te resulting regulations helped maintain the momen-tum, although they do not seem to have accelerated, theunderlying rate o improvement driven by the imperativeo economic eciency in a relatively ree market system,

and compounded by the transition rom a manuactur-ing economy to a service and knowledge based economy(Goklany a, pp. ). Consequently, envi-ronmental quality is much better now than in previousdecades. Carbon dioxide emissions, however, continueto grow. But this is due to the act that it is a late arrivalto societys list o environmental problems in act, itsimportance, given other global problems, is still con-tested and, in any case, theres been insucient time toaddress it economically (Lomborg ; Goklany ,a; Nordhaus ).

odays developing countries have been ollowing thepath laid down by the early developers. Many o themhave lower environmental quality than previously, but

7/27/2019 EJSD 2009

19/26

21

the electronic j ournal of sustainable development

www..g

because o the diusion o technology (which includesknowledge) rom developed countries, they are artheralong than early trailblazers such as the U.S. at the samelevel o economic development. For instance, in

when GDP per capita or low income countries was

$,, their lie expectancy was . years, a level thatthe U.S. rst reached in , when its GDP per capitawas $,. Even Sub-Saharan Arica, the worlds devel-opmental laggard, is today ahead o where the U.S. usedto be! In , its per capita GDP was at the same levelas the U.S. in but the U.S. did not reach Sub-Saha-ran Aricas current inant mortality level until , andlie expectancy until (estimated rom World Banka; Maddison , GGDC ; USBC ).

It can not be overemphasized that despite any envi-ronmental deterioration that may have occurred, the

well-being o the vast majority o the worlds humanpopulation has been improving continually over the pastseveral decades, as indicated not only by lie expectancy,but by other critical measures o well-being, including

poverty, mortality rates, ood supplies, education, childlabor, and so orth (Goklany a).

Why does reality not mirror Neo-Malthusianconcerns?

First, much o the environmental and Neo-Malthu-sian narrative implicitly or explicitly equates human

well-being with environmental well-being. While thelatter may be a component o the ormer, the two arentthe same. Few inside and even ewer outside rich coun-tries would rank environmental indicators among themost important indicators o human well-being, except

perhaps or access to sae water and sanitation. In act,the most critical indicators o human well-being lieexpectancy, mortality rates, prevalence o hunger andmalnutrition, literacy, education, child labor, or povertygenerally improved even during periods when otherenvironmental indicators were deteriorating (e.g., Figure

), indicating a lack o correlation between the two overthe long term. In act, long term trends are consistent

with the environmental transition hypothesis in that inits early stages, economic and technological develop-ment is negatively correlated with environmental quality,

whereas at high levels o development the correlation ispositive (Goklany a).

Second, as already emphasized, population growthhas slowed. It is no longer growing exponentially. Andauence and technology have much to do with that(Figure ).

Neo-Malthusians also overlook the act that in manyrespects auence, technology and human well-beingreinorce each other in what has been called the cycle o

progress (Goklany a, pp. ). I existing tech-nologies are not up to the task o reducing impacts orotherwise improving the quality o lie, it is possible with

wealth and human capital to improve existing technolo-gies or create new ones that will. HIV/AIDS is a case in

point.When HIV/AIDS appeared on the scene, it wastotally unanticipated. It was, or practical purposes, adeath sentence or those who contracted it. It took the

wealth and human capital o the most developed coun-tries to launch a response. Out o this came an under-standing o the disease and the development o varioustherapies. Once among the top ten killers in the U.S.,today HIV ranks nineteenth (counting all cancers andcardiovascular diseases as individual categories). From to , age-adjusted death rates due to HIV

declined by over percent (USBC ). Te richcountries have gured out how to cope with it, anddeveloping countries are beneting rom the technolo-gies that the ormer were able to develop because theyhad the necessary economic and human resources, andinstitutions at their disposal.

Tird, both technology and auence are necessarybecause while technology provides the methods toreduce environmental problems, auence provides themeans to aord them. In act, access to HIV therapiesin many developing countries is much higher becauseo wealthy charities and governments o the developedcountries (Goklany a, pp. ).

Fourth, there is a secular component to technologicalchange (see Figures and ), so that it ought to advanceeven i auence does not, provided we are open to scien-tic and technological inquiry. Tus, with secular tech-nological change and the mutually reinorcing advancesin economic development, the ability to reduce unto-

ward impacts and enhance the quality o lie has alsogrown rapidly.

Tese actors acting in concert over the long haul,have enabled technology or the most part to improveaster than either population or auence and helpedkeep environmental damage in check (e.g., or cropland)or even reverse it (e.g., or water pollution, and indoorand traditional outdoor air pollution), particularly in thericher countries (see able ).

able also shows that in the long run, technologyhas oen reduced impacts by an order o magnitude ormore. Tus, notwithstanding plausible arguments thattechnological change would eventually increase environ-

mental impacts, historical data suggest that, in act, tech-nological change ultimately reduces impacts, providedtechnology is not rejected or compromised via subsidies

7/27/2019 EJSD 2009

20/26

22

Have increases in popul ation, affluence and tecHnology worsened well-being?

Table 2: Changes in population, afuence and technology or various indicators

Indicator Area Period Population (P) Afuence(A

= GDP/P)

P x A = GDP Impact (I) Technology

actor (T)

Technological change

Total T in % T, in %/year

LAND (habitat converted to cropland)

cropland planted U.S. 19102006 3.22 6.24 20.08 1.00 0.050 95.0 3.1

cropland World 19502005 2.56 3.32 8.49 1.34 0.157 84.3 3.3

cropland India 19612005 2.43 3.19 7.77 1.05 0.136 86.4 4.4

cropland China 19612005 1.97 10.44 20.57 1.49 0.072 92.8 5.8

WATER WITHDRAWAL & USE

Water withdrawal U.S. 19502000 1.86 2.97 5.52 2.26 0.403 59.7 1.8

Water withdrawal World 19001995 3.16 6.27 19.8 6.54 2.07 107 0.08

Water consumption World 19001995 3.16 6.27 19.8 6.27 1.98 98.5 0.07

WATER (deaths due to water related diseases)*

Malaria U.S. 19001970 2.68 0.000 0.000 100

typhoid and

paratyphoid

U.S. 19001997 3.73 0.000 0.000 100

GI diseases U.S. 19001970 2.68 0.004 0.002 99.8 8.6

dysentery U.S. 19001998 3.78 0.014 0.004 99.6 5.5

AIR (indoor air pollution; residential emissions per occupied household)

SO2 U.S. 19402002 2.17 4.07 8.83 0.02 0.002 99.8 9.5

VOC U.S. 19402002 2.17 4.07 8.83 0.14 0.015 98.5 6.5

NOx U.S. 19402002 2.17 4.07 8.83 0.39 0.044 95.6 4.9

PM-10 U.S. 19402002 2.17 4.07 8.83 0.05 0.006 99.4 8.0

CO U.S. 19402002 2.17 4.07 8.83 0.05 0.006 99.4 7.9

AIR (national annual emissions)

SO2 U.S. 19002003 3.80 7.08 26.93 1.60 0.059 94.1 2.7

VOC U.S. 19002003 3.80 7.08 26.93 1.89 0.070 93.0 2.5NOx U.S. 19002003 3.80 7.08 26.93 7.94 0.295 70.5 1.2

PM-10 U.S. 19402002 2.17 4.07 8.83 0.29 0.033 96.7 5.4

CO U.S. 19402003 2.17 4.07 8.83 1.14 0.127 87.3 3.2

Lead U.S. 19702000 1.38 1.89 2.60 0.02 0.007 99.3 15.1

GLOBAL WARMING (extreme weather events, deaths, based on 10year averages)*

Deaths due to

climate-related

disasters

World 1900/09

1997/2006

3.67 0.17 0.047 95.3 3.1

Deaths rom

hurricanes

U.S. 1900/09

1997/2006

3.44 0.20 0.053 94.7 4.6

7/27/2019 EJSD 2009

21/26

23

the electronic j ournal of sustainable development

www..g

Indicator Area Period Population

(P)

Afuence(A

= GDP/P)

P x A = GDP Impact (I) Technology

actor (T)

Technological change

Total T in % T, in %/year

Deaths rom foods U.S. 1903/12

1997/2006

3.25 2.85 0.842 15.8 0.3

Deaths rom

tornados

U.S. 1917/26

1997/2006

2.65 0.02 0.075 92.5 4.1

Deaths rom

lightning

U.S. 1959/68

1997/2006

1.51 0.37 0.246 75.4 2.2

GLOBAL WARMING (carbon dioxide emissions rom combustion and industrial sources)

CO2 U.S. 19002004 3.84 7.28 27.91 9.12 0.327 67.3 1.1

CO2 U.S. 19502004 1.92 3.11 5.99 2.38 0.398 60.2 1.7

CO2 World 19002004 4.06 5.37 21.80 14.81 0.680 32.0 0.4

CO2 World 19502004 2.51 3.21 8.06 4.85 0.602 39.8 0.9CO2 China 19502004 2.37 11.74 27.81 63.66 2.29 128.9 1.5

CO2 China Since

economic

liberalization

19792004

1.34 5.06 6.76 3.31 0.49 51.1 1.3

CO2 India 19502004 2.97 3.63 10.77 20.16 1.87 87.1 1.2

CO2 India Since

economic

liberalization

19912004

1.24 1.73 2.16 1.84 0.85 14.8 1.1

*Death associated with these indicators are expected to increase with population but not with afuence (except through its eect ontechnology, which is captured in the T-actor). Thereore, the values o A and P x A are not relevant in these cases, and )T = percent

reduction in death rates over this period. Deaths due to malaria are rom USBC (1954) and Newman et al. (2004).

(which usually fow rom the general public to politicallyavored elements o society).

o summarize, population, auence and technolog yare not independent o each other. Moreover, technolog yis a unction o time. Tereore, in the IPA equation,the dependence o the I term on the P, A and terms is

not xed. It evolves over time. And the Neo-Malthusianmistake has been to assume that the relationship is xed,or i it is not, then it changes or the worse.

A corollary to this is that projections o uture impactsspanning a ew decades but which do not account ortechnological change as a unction o time and auence,more likely than not, will overestimate impacts, perhapsby orders o magnitude. In act, this is one reason whymany estimates o the uture impacts o climate changeare suspect, because most do not account or changesin adaptive capacity either due to secular technological

change or increases in economic development (IPCC, Figure SPM.; Goklany d; Reiter ;Southgate and Sohngen )).

8. Conclusion

Contrary to Neo-Malthusian ears, population is nolonger growing exponentially. Second, rom a historical

perspective, ood, energy and materials are more aord-able today than they have been or much o human

history. Tird, despite unprecedented growth in popula-tion, auence, consumption and technological change,human well-being has never been higher, and in the lastcentury it advanced whether trends in environmentalquality were up or down.

Tese outcomes were all possible because o greatereconomic and technological development, and, moreimportantly, the institutions that undergird such devel-opment (Goklany a). ogether, they steadilyimproved human well-being over the last century. Withrespect to the environment, however, their record is

mixed. Initially, in the rich countries, they exacerbatedenvironmental problems, but eventually they providedthe methods and means or cleaning up the environment.

7/27/2019 EJSD 2009

22/26

2

Have increases in popul ation, affluence and tecHnology worsened well-being?

Tat is, they went rom being part o the problem tobecoming part o the solution.

Developing countries, on the other hand, have yet tomake that transition or many environmental indicatorsin many places, although technological diusion, com-

bined with a little bit o auence, has allowed them tomove ahead o developed countries at equivalent levelso development.

In general, the world seems to have made the environ-mental transitions or access to sae water and sanitation,and lead in gasoline, and seems to be on the verge o atransition or cropland and water withdrawals.

So much or the past and present; what about theuture?

Humanity needs to improve the well-being o thebillions in developing countries that still suer rom

poverty and poverty-related problems such as hunger,malnutrition, contaminated water, malaria, and otherdiseases, while, over the next hal century, also accom-modating an additional three billion people and con-taining environmental impacts.

However, just as todays population couldnt be sus-tained and well-being improved with yesterdays tech-nology (able ), tomorrows population cannot besustained or its well-being advanced with todays tech-nology. Economic and technological development havebrought us this ar, and they are also necessary to moveus orward.

Without them, poverty, and all its consequences,cannot be reduced; the world will have to postpone thetransition or cropland; more land and water habitat

will be diverted to meet human needs; and we will be ina poorer position to cope with new and unanticipatedchallenges that the rest o nature may throw our way,including novel or resurgent diseases (such as anotherAIDS, or worse) or climatic changes.

But neither economic nor technological develop-

ment is guaranteed. Many policy preerences o someenvironmentalists and Neo-Malthusians, ounded ontheir skepticism o auence and technology, wouldonly make progress toward a better quality o lie anda more sustainable environment harder. Teir earscould become sel-ullling prophecies. Inklings o thiscan be seen in their antipathy toward genetically modi-ed crops which delays progress in reducing worldwidehunger and malnutrition even as it postpones an envi-ronmental transition or cropland and the developmento a more environmentally benign agriculture; in the

(ortunately) largely unsuccessul eorts to subordinatehuman well-being to environmental quality that wereused to justiy restrictions on the use o DD or public

health purposes; in their opposition to the developmento natural resources even under strict environmentalsupervision, which reduces supplies and increases prices;in the hostility to energy development whether it is ossiluels or nuclear, and which legitimizes dubious alterna-

tives such as land and water hungry biouels, solar arms,and dams; and at their dismay at the development oChina and India as they nally raise themselves rom a

poverty that the richer nations escaped rom a centuryago (Goklany a; Boqiang ).

Acknowledgments

I am grateul to Nicholas Eberstadt or bringing Paen-zeller () to my attention. I also thank two anony-

mous reerees, Julian Morris, and Caroline Boin, whosecomments and criticisms have vastly improved this

paper. Any remaining errors and shortcomings, however,are my responsibility.

I would also like to thank Luisa Molina or kindlysharing with me the data used in Figure , and CAI-Asiaor doing the same or the data used in Figure .

Notes

1. Te other two indicators are auence (as measured

by GDP per capita), and literacy (or some combined

measure or literacy and schooling o education).

2. Goklany (b) shows that a similar relationship holds

or cross country FR data or and using

GDP in constant dollars per market exchange rates

(MXR), whereas Figure uses PPP-adjusted GDP.

3. ime series analysis requently uses time as a crude proxy

or technology. See, e.g., Bhattarai and Hammig (:

), Shak (: ), Grossman and Krueger (:

), and Goklany (b). Te broader the denition

o technology, the better time serves as a proxy and, asnoted in the Introduction, technology is indeed dened

broadly in this paper because it includes both hardware

and soware (including knowledge, institutions and

rules o behavior).

4. ime series rom these sources were linearly extrapolated

to .

5. Figure uses the price o regular leaded gasoline rom

, the arithmetical average o regular leaded

and regular unleaded gasoline or , and

regular unleaded or . Nominal gasoline

prices are rom DOE (, ). All other economicdata are rom BEA (). Specically, the real price

o gasoline is calculated using the implicit GDP price

7/27/2019 EJSD 2009

23/26

2

the electronic j ournal of sustainable development

www..g

defator rom ables .. or - . Real disposable

income per capita is rom BEA (), able ..

6. However, it should be noted that the price defator used

in calculating the real price may not adequately take

into account the expansion o the money supply in the

past decade.

7. saw a very rapid increase in regular unleaded

gasoline prices during the rst hal o the year (rom

$. per (U.S.) gallon or the week ending January

to $. or the week ending on July) ollowed by an

unprecedented drop during the remainder o the year.

During the week ending January , the price was

$.. At its peak (during the week ending July , ),

the U.S. unleaded gasoline price was $. per gallon,

and the aordability index was at ..

8. Generally the impacts o an air pollutant are more

directly related to its concentration in the ambient air

rather than the total mass o its emissions. Tus, thepollutants outdoor (ambient) concentration, which

is measured in terms o the volume or mass o the

pollutant in a given volume o air (specied in terms

o parts per million, ppm, or micrograms per cubic

metre o air, respectively) is a much better indicator

o its public health impact than its gross emissions. In

recognition o this, the ambientair quality standardor any pollutant is almost universally specied in terms

o ppm or micrograms per cubic metre rather than

in terms o the mass o emissions. Consequently, it is

possible to improve air quality without reducing overall

emissions. Tis could be achieved, or example, through

better dispersal o the pollutant in the air through, or

instance, discharging exhaust gases containing emissions

into the air via higher chimneys or at high velocities or

high temperatures.

9. Mortality rate is the number o people dying due to

extreme weather events divided by the total population

exposed to the events.

10. Te improvements due to increasing economic eciency

and the shi rom manuacturing to service and

knowledge based economies are also refected in Figures

through .11. Te UNDPs Human Development Index, or instance,

is based on three indicators: lie expectancy, per capita

income and some combined measure o education and

literacy (UNDP ).

12. Tis also helps explain why these environmental

problems are among the rst to be solved, and why

lack o such access trends downwards with economic

development (Shak ), indicating that virtually

every country has gone past its environmental transition

or these indicators (Goklany a).

Reerences

Ausubel, J.H. . Does Climate Still Matter?Nature :.

Ausubel, J.H. . Resources and Environment in the

st Century: Seeing Past the Phantoms,World EnergyCouncil Journal(July): .

[BEA]. Bureau o Economic Aairs. .NationalEconomic Accounts. Available at http://www.bea.gov/bea/dn/nipaweb/Selectable.asp. Visited January .

Blake, E.S., Rappaport, E.N., Landsea, C.W., and Miami

NHC . Te Deadliest, Costliest, and Most IntenseUnited States Hurricanes om 1851 to 2006 (and other

Frequently Requested Hurricane Facts). NOAA echnical

Memorandum NWS PC-.

Boqiang, L. . We cant aord to pollute rst, improve

later. China View: June . Available at http://news.

xinhuanet.com/english///content_.htm . Visited on July .

Bruce, N., Perez-Padilla, R. and Albalak, R . .

Indoor air pollution in developing countries: a major

environmental and public health challenge, Bulletin o

the World Health Organization : .

[CAI-Asia] Clean Air Initiative or Asian Cities. .AirQuality in Asian Cities. On le with author.

Commoner, B. Te environmental cost o economic

growth, Chemistry in Britain : .Dahl, R. . Te Population Equation: balancing what

we have with what we need.Environmental HealthPerspective: : A-A.Dahl, .E. . Status and trends o wetlands in the

conterminous United States to . Washington,

DC: U.S. Department o the Interior, Fish and Wildlie

Service. http://wetlandsws.er.usgs.gov/status_trends/.

Diamond, J. . Collapse: How Societies Choose to Fail orSucceed(New York: Viking).

Department o Energy. .Motor Gasoline Retail Prices,U.S. City Average. Available at http://www.eia.doe.gov/emeu/mer/prices.html.

Department o Energy. . Weekly Retail Gasoline and

Diesel Prices. Available at http://tonto.eia.doe.gov/dnav/pet/pet_pri_gnd_a_epmr_pte_cpgal_a.htm. Visited

January .

Dumitrescu, E. . Global Phaseout o Leaded Gasoline.United Nations Environment Program.

Economic Research Service. .Major Land Uses,1910

2006. Available at http://www.ers.usda.gov/Data/

MajorLandUses/spreadsheets/croplandusedor