It was devine nature which gave us the country, and man's skill that built the cities

- Marcus Terentius Varro, De Re Rustica

Among the greatest distinctions of the twentieth century were its powerful twin surges of population growth and urbanization. These trends reflect billions of human choices, conscious and nconscious, made for countless reasons. Some of these choices were individula, such as weither to marry or where to live. Some were political, such as the Australian choice to build a capitol city at Canberra in the 1920s and the decision of South Africa's National Party to implement the apartheid in the 1940s. All these decisions collectively generated the global treends of population growth and urbanization. They affect- ed everything in human affairs, to greater or lesser degrees, from high culture to child nuture to corporate structure. They effected much that was not human too.

Something New Under The Sun P 270

Population Growth

Most discussion of the social forces behind environmental change is politically charged, and none more so than the issue of population. De bate often boils down to arguments that other people must change their ways to save the earth: Indians and Africans usually argue that population growth matters little; Americans and Europeans that it matters greatly. My view is that it mattered for some varieties of environmental change but not others, and that migration was often more consequential than sheer growth. The issue is anything but straightforward.

The bizarre population history of the twentieth century was the climax (to date) of a long frenzy of reproduction and survival. Chapter i consid- ered the very long term; here I will focus on the past 500 years. The glob- alization brought about by European mariners at the end of the fifteenth century incubated two biological changes significant for subsequent pop ulation history. First, their globe-girdling voyages spread diseases among populations that had previously been long isolated from one another, li the short run, this led to catastrophic losses, notably in the Americas and Oceania. Eventually this swirl of infection produced more seasoned im- mune systems, closer symbiosis between pathogens and hosts, and still later, public health systems, so that the toll from epidemics subsided. Secondly, maritime voyagers distributed food crops far and wide, even- tually allowing (within limits imposed by politics and tradition) every re- gion of the world to concentrate on the crops best suited to its ecologica and market conditions. Maize spread from its home in the American tropics to East Asia, southern Africa, and the Mediterranean basin. Trop ical Africa acquired manioc (cassava) from Brazil, while the Americas got wheat. In sum, the worlds food supply improved. By about 1650 these two factors provoked a long-term rise in world population, still in train. On top of this surge, itself unprecedented in human history, came tidal wave of population growth. It drew on further improvements of food supply and disease prevention (some described in Chapter 7) and reached a crescendo after 1950. These improvements first caused de- dines in mortality, to which some societies later responded by restricting their fertility. In between, when mortality had declined but fertility had not, population grew very quickly. Demographers call the whole process

More People, Bigger Cities 271

"the demographic transition." (Note 1) The rate of growth peaked in the late 196o's (at 2.1% per annum). It so appalled observers that a careful and sensitive one, Kenneth Boulding, seriously suggested establishing trad- able permits for having children. (Note 2) Growth slowly slackened after 1970, mainly because women acquired more say in many societies and limited their fertility. By 1996 the total annual increment of population had peaked at about 92 million to 95 million more births than deaths per year. Population growth quickened and slowed at different times in differ- ent places. The demographic transition began first in Europe, where it took more than a century to complete. In East Asia it came only after 1950. but took less than half a century. In Africa, it remained unfinished, for fertility by the late 199os had only recently and spottily begun to sub- side. Table 9.1 gives a rough idea of the pace of population growth by world region.

In the period 1850 to 1950 the populations of Africa, Asia, and Europe roughly doubled. Meanwhile numbers in the Americas, Australia, and

Note: The figures for Africa and to a lesser extent Asia are speculative prior to 1900; figures some-what lower for Africa and somewhat higher for Asia appear in McEvedy and Jones 1978.

Note 1:
The transition brings birth and death rates down from around 30-35 per thousand (per year) to around 10-12 per thousand.

Note 2:
Boulding 1964: 135-6. Boulding (1910-1993), an economist and Quaker, was not normally given to recommending sharp restrictions on freedoms.

P. 272

Oceania grew much faster, five- or sixfold in ioo years. This reflected both migration patterns and differences in natural increase. After 1950 the locus of fast growth changed. In the ensuing half century, Asian num- bers more than doubled, Latin American population tripled, and African population nearly quadrupled. Meanwhile Europe and North America grew much more slowly, having completed the demographic transition by 1950. (Note 3)

By the 1990s humankind accounted for about o.o1 percent of earths total biomass, and about 5 percent of its animal biomass, ranking with cat- tle and far outstripping any other mammal. (Note 4) This crowning success of the human species coincided with intense environmental change. Did it cause that change? I will try to answer this with both arithmetic and anec- dote. First the arithmetic.

Consider the relationship between global population growth and some global air pollutants. Between 1800 and 1990, world population increased by a factor of 3-5, while emissions of carbon dioxide, the main green- house gas, climbed more than 17-fold. (Note 5) A superficial calculation (note 6) suggests that population then accounted for only about a fifth of that rise. In the same span, global emissions of sulfur dioxide, a major component of acid rain, increased about i3-fold. Population increase accounted for slightly more than a quarter of the growth in sulfur emissions. If one does the arithmetic for the United States alone, population growth ac- counted for 31 percent of the growth in carbon dioxide emissions; in what is now OECD (that is, the rich countries of) Europe, 41 percent; iri Japan and the USSR (and successor states), only 2 percent; in Africa, less than 1 percent. (Note 7) Abandoning this primitive arithmetic for the moment,

Note 3:
The data would look different if presented not by continent but by religion or GNP per capita. However, these indices have changed so much in the 20th century that it is impractical to do so.

Note 4:
Turco 1997:105· Ants outweigh us about 4:1.

Note 5:
These data are from RIVM 1997.

Note 6:
The calculation is superficial for several reasons. It deals with global aggregates, and neglects the possibility that population growth occurred in places where carbon dioxide emissions did not and that the two were unrelated. It assumes a linear relationship between the two, whereas in reality there might well be thresholds below which population increase makes little difference, or above which it does. Regrettably, no precise mathematical relationship between these variables can logically be constructed.

Note 7:
The reason these figures vary so much is that in western Europe and the USA, population growth after 1890 was slow (by world standards) and carbon dioxide emissions were already (comparatively) high. In Japan and Russia, while population growth was also comparatively slow, carbon emissions were very low in 1890 because the countries had just begun to indus-

MORE PEOPLE, BIGGer CITIES 273

one may safely suppose that population growth had a minimal role in re- leasing chlorofluorocarbons (CFCs) into the stratosphere.

For some important forms of air pollution, then, population growth in the twentieth century was a significant but by no means preponderant driving force. This stands to reason: combustion generated most of the air pollution, and the quantity of combustion and the intensity of its pollu- tion were only loosely linked to the number of people. In rich societies, such as the United States and Germany, additional people did substan- tially raise air pollution levels from 1900 to 1970 or so, because they drove cars, heated with oil or coal, and in general increased total com- bustion. After 1970, new technology, encouraged by regulation, lawsuits, and a decade of high energy prices, loosened the link: cleaner production and cleaner cars meant additional people caused much smaller incre- ments in air pollution than they had in the 1950s or i96os. In poor soci- eties additional people had even less impact on air pollution because they caused negligible increases in combustion. Even where population growth and pollution coincided, as in China after 1970, it is hard to con- clude that the former caused the latter. Fast and careless industrialization and urbanization probably mattered more than the rate of population growth. (Note 8)

In general, population growth provoked additional pollution of air and water primarily where and when the economy was already industrialized

Note 7 continued:
trialize. Harrison (1992) does similar arithmetic for 1961-1988 and concludes that populatior accounted for 44% of the growth in carbon emissions. Taking these years, when populatior growth was at a maximum and when emission increase rates declined (after 1975) accounts for the higher figure. Ogawa 1991 and Darmstadter and Fri 1992 examine the years 1973-1987 and rate the role of population growth even higher. In those years, improved energy efficiencv and reduced carbon emissions per unit of energy consumption combined to lower carbon emissions by 1% per year, but this was more than counteracted by the impact of population (+1.8% per year) and economic growth (+1% per year). These 15 years were ones of very fast population growth and (by post-1945 standards at least) slow economic growth, so, like Harrisons figures, they are helpful on their own terms but not useful for the century as a whole. Other arithmetic exercises along these lines include Raskin (1995), who finds population growth responsible for 32% of carbon dioxide emission growth for the period 1950-1990; a fine summary of the weaknesses of all such calculations is found in McKellar et al 1998:120-35.

Note 8:
China's extraordinary demographic trends are summarized in Lee and Feng 1999. Population policies since 1978 kept Chinese population 250 million below what it otherwise would have been in 1998. Growth rates in the 1960s approached 3% per annum, but in the 1970s dropped to under 2%. In 1996, Chinas annual population growth rate was 1.%. Smil 1993considers pollution and population in the recent Chinese context.

P 274

and where society (and state) did not value environmental amenities. This was true in the United States, Japan, and western Europe from roughly 1890 to 1970, in Russia from 1060 onward. Population growth in societies without significant industry had much less impact on pollution levels except for human wastes and domestic smoke. In societies under- going industrialization, such as South Korea (1960-1990) or the USSR, (1930--1960), the rate of population growth mattered much less than the rate and type of industrialization.

The nexus between population and pollution in the twentieth century is hazy enough. But the relationships between population growth and other forms of environmental change are cloaked in still thicker confu- sion. Population pressure both caused and prevented soil erosion. In places where it drove farmers up steep hillsides, as in Java or northern Morocco, it quickened erosion. Elsewhere it provided enough labor to build and maintain soil conservation schemes, as in the Machakos Hills of Kenya. Moreover, in mountain environments the loss of population sometimes brought on faster soil erosion, as too few people remained to maintain terraces and other soil conservation stratagems. Soil salinization sometimes derived (indirectly via expanded irrigation) from population pressure, but more often from the temptations of commercial or centrally planned agriculture. Population growth and density were only partial de- terminants in these equations: natural, political, and economic condi- tions frequently carried greater weight. The best conclusion-a rough one-is that population growth often heightened erosion rates, but dense populations, if stable, could lower them. (Note 9)

Population growth probably accounted for much of the worlds in- creased water use and intensified problems of water scarcity (see Chap- ters 5 and 6). A superficial calculation suggests 44 percent of it: water use increased ninefold and population fourfold in the years 1900 to 1990, so four-ninths of the century's increment derived from the existence of more people. This, however, is only a rough measure. Changes in water-use ef- ficiency, as well as in pricing and subsidies, blur the picture. In the United States after 1980, as noted in Chapter 5, water use declined while popu-

Note 9:
On Java, see Repetto 1986, who reports a sixfold increase in erosion rates (1911-1983) attributable to population pressure. On northern Morocco, see Maurer 1968; on labor shortage and erosion, see Barker 1995' McNeill 1992b, and Mignon 1981. The tale of the Machakoi Hills was given in Chapter 2. Interestingly, in the 19305, Jacks and Whyte 1939:286-7 argued that a dense population was the best insurance against soil erosion, offering Japan and Java evidence (and admitting India and China as exceptions)!

p 275

lation grew. In almost every society considerable slack existed in the water-use systems, such as inefficiencies and waste, so changes in tech- nology and policy could, and sometimes did, alter matters more sharply than could population growth. (Note 10)

Population growth surely played a large role in driving the manifold changes to the biota in the twentieth century. Food demand drove most of the centurys doubling of cropland, helped fuel the Green Revolution, and multiplied the worlds fishing effort. Population growth did not pro- duce these changes alone, but in matters directly related to food pro- duction it loomed largest.

Some of the important biotic changes, however, had little connection to population or demand for food. Whaling, unlike fishing, did not sig- nificantly reflect expanding food needs. Biological invasions had almost nothing to do with human population growth. The vast shifts in human- microbial relations had a lot to do with it, but here the causation was re- versed: the environmental shift drove population growth.

Deforestation admirably illustrates the murky conundrum of environ- ment and population. In some cases, such as rural Ethiopia, recent stud- ies conclude that population growth was the main driving force. But historical studies extending back to the nineteenth century show scant forest even when Ethiopia's rural population was only a fraction of what it became and when its growth rate was also much lower. Deforestation, in Ethiopia and around the world, occurred in conditions of population growth, population stagnation, and even population decline (e.g., Russia in the 1990s or Madagascar in the period 1900-1940). (Note 11) A meta-analysis (that is, a statistical study of numerous independent studies) of the rela- tionship between population and deforestation, concluded that

"while population pressure is an important force leading to deforestation, it rarely acts alone to produce the outcome. Other determinants appear to be necessary' as mediation and contingencies for population growth (or

Note 10:
See Falkenmark 1996 for a discussion of contemporary population and water issues. She seems to regard population as a stronger driving force than my rough calculation suggests. The discrepancy, if it is that, may derive from the fact that population growth in the 1990s (Falkenmark's implicit reference point) is greater than the average for the 20th century (my reference point).

Note 11:
On Ethiopia, see Campbell 1996, Grepperud 1996, and McCann 1997. Madagascar lost 4 million ha (hectares?) of forest (1900-1940), while population stagnated or possibly declined. Cash crops, notably coffee, replaced the forest (Jarosz 1993. Kummer 1991:146-9 concludes that population growth played a scant role in deforestation in the Philippines in recent decades.

P 276

density) to have a discernible impact. . . . [Q]uantitative analysis suggested that even if the effects of population growth are statistically significant, their magnitude is quite modest." (Note 12)

Such a vague conclusion, regrettably, is just about right.

In sum, population growth accounted for a modest share of air pollu- tion-related environmental changes and a larger share of those pertain- ing to water and the biota, especially those involving food production. Big environmental change resulted more often from combinations of mutu- ally reinforcing factors than from population growth alone. The latter probably mattered more late in the century than earlier, mainly because growth climaxed after 1960. (Note 13)

Migration
--------------------------------------------------------

Migration often mattered even more than growth, although the two are often hard to separate because growth sometimes caused, or at least helped cause, migration. From 1500 to about 1870 most of the worlds in- tercontinental migrants were slaves or "coolies." Then, around 1845 to 1920 the spontaneous migration from Europe to the Americas came to overshadow other currents around the world. After 1925, international migration receded for several decades, and when it gathered pace again by 1960, more diffuse flows prevailed. In all periods, migration often promoted radical changes in land use and sharp ecological shifts.

From the point of view of environmental change, the most important migration involved frontier areas. Mass migrations from humid to dry lands repeatedly provoked desertification. (Note 14) Migrations from flatlands to sloping lands often led to faster soil erosion. Migrations into forest zones brought deforestation.

Note 12:
Palloni 1994:160.

Note 13:
Repetto and Holmes 1983 attempt to show mathematically the limited role of population in accounting for natural resource depletion (using mostly 19705 data). The view that population growth is "probably predominant. . . in environmental problems" is put forward by Myers 1993.

Note 14:
Here desertification means progressive loss of vegetation cover and organic content in the soils. The process is not necessarily irreversible.

P 277

With declining transport costs and tightening integration of markets after 1870, people moved about as never before. They frequently ven- tured into ecological zones about which they understood little. This, of course, has happened throughout human history. But in the twentieth century more people moved, they had more transformative technologies at hand, and in most cases they had stronger links to markets or national planners, giving them incentives to clear more land, plant more crops, graze more animals, catch more fish, or mine more ore than their own subsistence required.

The scale was enormous. In the period 1830 to 1920, Europe alone sent 55 million to 70 million emigrants to the Americas, Australia, and Siberia. 5 A large minority engaged in pioneer farming, such as Ukraini- ans on the Canadian prairies or Italians on the coffee frontier of Brazil. (In 1934 nearly half of Sao Paulo state's coffee farms were owned by im- migrants). (note 16) This epochal migration wound down after its 1913 peak, restricted by World War I, anti-immigration laws in the United States (1924) and elsewhere, anti-emigration laws in the Soviet Union (1926) and elsewhere, and after 1929 by the Great Depression. Simultaneously (1834-1937) about 30 million to 45 million Indians moved, mostly as in- dentured laborers to plantations in Fiji, Malaya, Burma, Mauritius, Natal, Trinidad, and Guyana. While they took no part in land-use decisions, and while most were sojoumers not settlers, their labor converted millions of hectares of forest land to sugar, rubber, and other crops. Large numbers of Chinese labor migrants moved to Southeast Asia, the Caribbean, Cal- ifornia, and Peru among other destinations. This "coolie" trade too slowed under the impact of war and depression after 1914. (Note 17)

But migration did not stop. It became less international. Millions upon millions of people after World War I migrated into new rural areas. Fewer and fewer people did so spontaneously. Instead, compulsion and state policy took a larger role after 1920, a return to the pattern that prevailed from 1500 to 1870. In some cases, state policy promoted voluntaiy mi- gration by offering financial and other incentives. Great Britain paid young people, especially war veterans, to migrate within the Empire to Australia, New Zealand, and Canada. A few hundred thousand signed up, and a few million hectares in the Dominions were consequently broken to the plow. In other cases, state policy forced migrations, as in South

Note 15:
Baines 1995; Thumerelle 1996:106-7

Note 16:
Klein 1995:211

Note 17:
Zolberg 1997:288

P 278

Africa. There the Land Act (1913) and subsequent laws restricted own- ership of good farmland to whites, obliging millions of black South Africans to move to ostensible tribal homelands called bantustans. There they crowded onto South Africa's poorer lands, farming steep slopes and running their cattle on semiarid bushlands, resulting in an archipelago of intense land degradation in that country. Mao's plans after 1949 moved millions of Chinese, accustomed to monsoon agriculture, to dry lands in Inner Mongolia, Xinijiang, and Tibet. Brazils rulers made settlement of Amazonia a national goal after 1960, calling it "a land without men for men without land." In all these cases, mass migration served as a means to political or social ends. Ensuing environmental changes were some- times unforeseen side effects; at other times they were expected, ac- cepted, or even desired by the leaders in question. Two of the more determined state efforts, involving varying proportions of compulsion and incentives, took place in the Soviet Union and Indonesia.

From at least the sixteenth century Russians had settled lands beyond the historic Russian heartland. The search for land and furs brought them to Siberia, central Asia, and Alaska. This epic of national expansion con- tinued, except in Alaska, through the end of the tsarist period in 1917. In the final decades of Imperial Russia, the trans-Siberian Railroad brought 3 million to 4 million Russian peasants in search of free land in Siberia. The government, eager to develop Siberia, encouraged this migration and augmented it with a flow of political exiles. (Note 18) Soon after the formation of the USSR (1917-1922), state-sponsored migration quickened. Soviet settlement in Siberia and Central Asia served two purposes. It was both a punishment for counterrevolutionary individuals and nationali- ties, and part of the brave new future of economic development planned by Soviet leaders. Mass migration began in 1929 with Stalin's collec- tivization of agriculture. Between 2 million and 3 million Russians and Ukrainians were forcibly removed to Siberia, the Urals, or the Russian far north. Over the whole Stalinist period (1927-1953) more than 10 million people suffered forced migration. More than half were sent beyond the borders of the Russian republic, often to "virgin land." During World War II Stalin deported entire peoples whose loyalty he distrusted. One million ethnic Germans and hundreds of thousands of Chechens, Tatars, and Kalmyks moved to Central Asia and Siberia in the desperate and

Note 18:
Treadgold 1957-159-61. Chesnais 1995:221 gives 5.8 million as the total number of migrants for 1801-1914, from European Russia across the Urals. Of these, 80% were voluntary and 20% were internal exiles.

P 279

chaotic days of 1941 to 1944. Millions of Russians took over their lands along the lower Volga, in the North Caucasus, and Crimea.

Stalin was a breaker of nations, and the swirl of migrations he ordered moved millions into unfamiliar environments. Some hacked away at northern forests, others dug canals, built railroads or industrial com- plexes, while still others tried to till the arid steppe. After Stalins death in 1953 the Soviet Union used the carrot more than the stick to move an- other million Russians to plow the Kazakh steppe in Khrushchev's Virgin Lands program (see Chapter 7). All in all, several million Soviet citizens found themselves toiling away in unfamiliar landscapes, radically altering land use and cover, hydrology, soils, and much else. Their government in- tended that they should do so, for the glory of socialism and the strength- ening of the USSR. Given their struggle to survive, it is safe to suppose notions of soil conservation and forest preservation rarely crossed their minds. (Note 19)

Indonesia, like the Soviet Union, incorporated vast expanses that its rulers sought to transform into economically productive areas. It used less compulsion and moved fewer people. The Indonesian scheme was known as "transmigration."

Most Indonesians live on Java, which together with Bali makes up the agriculturally fertile inner core of the country. The far larger outer islands of Borneo, of which the Indonesian portion is called Kalimantan, and Sumatra, together with thousands of smaller islands, have among the poorest soils on earth. Java and Bali have long hosted dense populations of rice farmers, who over the centuries carved their islands' rugged ter- rain into mosaics of irrigated terraces.~ The outer islands, in contrast, supported scant population, remained mostly under forest, and con- tributed little to the wealth and power of the state. That state for cen- turies was a colonial one, run by the Dutch.

The Dutch began the transmigration project in 1905, acting on ideas in circulation since at least the eighteenth century. Dutch colonials had long thought that the population of Borneo and Sumatra needed augmenting so that the resources of these islands, notably timber and gold, could be brought to market. (Note 21) In 1905 the Dutch East Indies had about 37 million

Note 19:
The migration data come from Polyakov and Ushkalov 1995. Zemskov 1991 reports the forcible migration of 1.8 million kulaks in 1930-1931 alone. A zestful account of logging's ravages in the 1930s is reported in Andreev-Khomiakov 1997:27-38 et passim.

Note 20:
See Geertz 1963.

Note 21:
Knapen 1998.

P 280

people, of whom 30 million lived on Java. By awarding nearly a hectare of land to any family that would move, the Dutch induced about 200,000 people to migrate, mostly to southern Sumatra, by 1941. This did next to nothing to relieve population pressure, a stated goal. In its peak year (1941) Dutch transmigration involved 36,000 people. Java acquired this many people every two to three weeks from natural population increase. Transmigration stopped while Japan occupied the islands (1942-1945) and resumed when Indonesia became independent in 1949. Indonesian planners hoped for an outflow from Java of 2 million people per year, and intended that by 1987 no less than 50 million Javanese would populate the outer islands. This would alleviate pressures on Java, turn the outer islands into revenue-producing areas, and conveniently swamp their local popu- lations, whose devotion to the Indonesian state at best wavered. This grand scheme enjoyed the support of diverse backers, from the Indone- sian Communist Party to the World Bank, which chipped in half a billion dollars. But it foundered on-among other things-ecological ignorance.

Reality proved triply disappointing. Too few took the lure, most who did stayed poor, and the resentments created by transmigration threatened state security more than the process of swamping indigenous people with loyal Javanese helped. By 1987 only 4.6 million people had moved, equiv- alent to about three years' increment in Javas population growth at 1970s rates. The lure had grown to 4 to 5 hectares of land per family. But still the vast majority of Javanese preferred the crowded conditions of rural Java, the squalor of Djakarta's shantytowns, or labor migration to Malaysia to the hardships of pioneer life on the outer islands. Javas peasants were experts at growing paddy rice, but their skills applied in only a few places in the outer islands. Their heroic efforts to clear the forest too often led to poor harvests, dashed hopes, land impoverishment, and abandonment. On the outer islands luxuriant forest growth derived from constant nu- trient cycling between trees and the organic litter on the forest floor. The soils contained few nutrients, as is widely the case in the tropics. Farm- ing successfully in such environments required different skills, normally swidden techniques. (Note 22) Some transmigrants adopted appropriate tech- niques, some did not. Some enjoyed strong state support in the form of

Note 22:
Swidden or slash-and-burn is a technique in which farmers burn forest, plant crops among the ashes, harvest their crops for one to three years, and then move on to repeat the cycle elsewhere. They (or others) can return to the original site and use it again, but only after 10-30 years, enough time to allow sufficient vegetation to grow that, when burned, will release enough nutrients for farming.

P 281

schools, health clinics, and so on, but others did not. In many cases mi- grants gave up on eroding lands, on fields invaded by tough and useless imperata grass and other pests. Many found themselves in conflict with local populations on the outer islands, who wanted to farm the best lands themselves. About 20 percent of the migrants improved their standard of living; many of the rest felt swindled. By the late 1980s the transmigra- tion scheme had ground nearly to a halt. (Note 23) While it lasted, it provoked considerable environmental change; had it succeeded as envisioned by its promoters, it would have changed the outer islands utterly.

Officially sponsored migration helped decimate two of the twentieth century's three great rainforests, in Indonesia and Brazil. Africa's rain- forests avoided this fate in part because of the continents political frag- mentation: impoverished Sahelian farmers and herders were not encouraged to migrate to Zaires forests. Had Africa been unified, as pan- •I- Africanists had hoped, the life expectancy of Africa's tropical forests would surely have been shortened. (Note 24)

Around the world, spontaneous and officially sponsored migrations moved tens of millions of people into ecologically unfamiliar areas in the twentieth century. A hefty proportion of the worlds environmental changes resulted. This was especially true in those categories strongly affected by frontier farming: vegetation cover, biodiversity, soil condi- tion, and, in dry lands, water use.

The Footprints and Metabolisms of Cities
--------------------------------------------------------------------

Twentieth-century urbanization affected almost everything in human affairs and constituted a vast break with past centuries. Nowhere had humankind altered the environment more than in cities, but their im-

Note 23:
Abdoellah 1996; HardJono 1977, 1988; Levang and Sevin 1989. The World Bank's role is treated in Rich 1994:34-8.

Note 24:
Note the limits of this argument. Tropical deforestation in Brazil and Indonesia was caused by other things as well as in-migration (see Chapter 8); in-migration in these countries took place spontaneously as well as with official help; and international migration within Africa, botth during and after the colonial period, did (to a small degree) transform forests into croplands most notably in West Africa. On the migrations to the West African forest zone, see Cordell et al. 1996.

p 282

pact reached far beyond their boundaries. The growth of cities was a cru- cial source of environmental change.

First, consider the scale and tempo of the process. A millennium ago China and the Islamic Middle East had the worlds most urbanized pop- ulations, but even in these lands 90 to 95 percent of people lived outside cities. City size was strongly constrained by the nearly stagnant agricul- tural productivity of their hinterlands and the sharply fluctuating ability of cities to dominate these hinterlands politically. In 1700, only five cities in the world, all political capitals, had more than half a million people: Is- tanbul, Tokyo, Beijing, Paris, and London. By 1800 there were still only six. (Canton/Guangdong had joined the list.) The hazards of city life, mainly endemic childhood diseases but also epidemics, killed people faster than others were being born. London, for example, in 1650 re- quired 6,000 in-migrants per year to stay even. In 1750 Londons lethal- ity counteracted half the natural increase of England as a whole. (Note 25) But the constraints on urban growth soon relaxed: farms were growing slightly more productive; cities, at least capital cities, were more able to dominate hinterlands; fossil fuel transport extended the reach of cities farther into the countryside; and public health programs began to take effect. By the late 1880s, city folk in Austria and Bavaria had longer life expectancies than their country cousins. By the 1920s, urban Chinese outlived Chinese peasants. (Note 26) For 8,000 years cities had been demographic black holes. In the span of one human generation they stopped checking population growth and started adding to it: a great turning point in the human con- dition.

By 1900 there were 43 cities of over half a million, mostly in western Europe, eastern North America, and on the seacoasts of export-oriented economies, often in European colonies. By 1990 about 800 cities, scat- tered all over the world, surpassed the half million mark. Some 270 had more than a million residents, and 14 topped 10 million. It remains to be seen what constraints on city size now exist. (Note 27)

The first country in which urban dwellers accounted for a tenth of the population was the seventeenth-century Netherlands, propelled by the

Note 25:
Macfarlane 1997:22.

Note 26:
On Austria and Bavaria, see Hohenberg and Lees 1985:259, and Munch 1993. On China, see Lee and Feng 1999. This transition may have begun first in Japan (Hanley 1987). One the few exceptions to the greater longevity of urban populations in the 20th century was 1980s Poland, where acute pollution contributed to shorter lives among city folk.

note 27:
Berry 1990:104-6; WRI 1007:8-9.

P 283

twin engines of global commerce and unusually rich local agriculture. The first country in which half the population lived in cities was industrializing England, in about 1850. The United States reached this level at about 1920, Japan about 1935, the USSR and Mexico about 1960, South Korea about 1975, South Africa about 1985. In 1998 the world as awhole approached it. (Note 28) In very general terms, rapid urbanization happened in Japan, western Europe, and eastern North America in the nineteenth and early twentieth cen- tury, in the Soviet Union and Latin America in the mid-twentieth century, and almost everywhere else after 1960. In national terms, by far the fastest urbanizations occurred in the USSR in the 1930s and in China in the 1980s. Each of these reflected burgeoning industrialization, in the Soviet case while building communism (and starting cities from scratch) and in the Chinese case while dismantling it. Chinese policy under Mao had success- fully stalled urban growth, but after his successors abandoned Maos re- strictions in the late 1970s, cities grew with pent-up force. Table 9.2

Note: These data ultimately derive from national censuses (and estimates) that define "city" and "urban" very' differently. In Japan a community normally had to have 20,000 people to qualify as urban; in Turkey, 10,000; in the USA, only 2,500.

Note 28
By one calculation (Clark 1998), the world was already half urban by 1996.

P 287

summarizes the recent history of urbanization for some large regions of the world. As the table shows, the worlds urban percentage tripled in the twen- tieth century. The total number of urban dwellers rose from about 225 mil- lion in 1900 (most of them in Europe and North America) to 2.8 billion in 1998, a 13-fold rise.

The environmental meaning of the worlds tumultuous urbanization was vast and variable. Urban impacts extended beyond city limits to hin- terlands, to downwind and downstream communities, and in some re- spects to the whole globe. Cities absorbed ever larger quantities of water, energy, and materials from near and far. In exchange they pumped out goods and services-as well as pollutants, garbage, and solid wastes. Broadly speaking, this process of urban metabolism generated two cate- gories of environmental change: pollution effects and land use effects. (Note 29)

GARBAGE AND POLLUTION. Rapid urbanization normally generated severe pollution stress. Infrastructure rarely kept pace with heady urban growth. Many people in 1900 and about 800 million in 1990 lived with- out piped water, sewage systems, gas, or electricity. (Note 30) Consequently they lived amid their own wastes and pollution (see Chapters 5 and 6). No novelty in that: most city dwellers from ancient times had lived-and died- that way. But far more participated in this grim existence in the twentieth century, often in ramshackle shantytowns built around urban cores. Shan- tytown inhabitants usually lacked formal title to their homes, could be evicted at any time, and consequently invested little time or money in neighborhood improvements. City governments occasionally did so, as in Curitiba (Brazil) during Mayor Jaime Lemers tenure in the 1970s and 1980s, when recycling and public transport attained unusual efficiency. But often they did not, because of lack of funds or lack of concern. The environmental conditions created by speedy urbanization were among the bleakest humans experienced. Slimy water and gritty air were the worst but not the end of it.

Consider garbage and solid waste. In the nineteenth century, cities everywhere-except maybe Japan-reeked of garbage. (Note 31) But when ad- dressing water supply and sewage issues after 1800 or so, most munici- palities in the Western world also organized garbage collection and

Note 29:
General treatments with some historical dimension include Berry 1990, Douglas 1994, and Gugler 1996.

Note 30:
WRI 1997: 52-3 has a table with these data for 1980-1995.

Note 31:
Hanley 1087 says Japanese cities organized garbage collection in the seventeenth century.

P 288

disposal. Early in the twentieth century, New York had garbage barges that dumped their contents in the waters outside the city's harbor. Other cities did things differently, but almost all shunted garbage onto neigh- boring land or into neighboring bodies of water, improving urban health. By midcentury the scale of garbage and solid-waste problems began to grow, overwhelming the capacity of many cities to cope. Mexico City in 1950 produced about 3,000 tons of garbage per day. By the late 1990s it generated three times as much. At least a quarter of it piled up in streets, ditches, and ravines. Garbage hills and ridges sprouted in almost every fast-growing city, from Manila to Maputo and Quito to Karachi. Streets and alleys acquired a rancid and rotting patina of household and other wastes. In Surat, a city of 2.2 million in India, a fifth of the citys garbage went uncollected in the early 1990s··· This created ideal rat habitat, and helped launch an outbreak of bubonic plague in 1994. Prompt official re- sponse and tetracycline limited the toll to 56 deaths. Galvanized by the plague, Surat managed to clean its streets and collect its rubbish so effi- ciently that by 1997 it was deemed India's second-cleanest city. (Note 32) Few other cities, however, matched this civic miracle. Fortunately, highly un- sanitary conditions have to date produced no sweeping epidemics of the sort that culled urban population in earlier centuries, testimony perhaps to the power of modern medicine.

Where infrastructure caught up to urbanization, solid waste was col- lected and kept away from human population. After World War II, New York City consigned its trash to Staten Island, where the worlds largest, landfill opened in 1948, now a towering monument to wastrel ways. By the late 1990s, New York sent caravans of garbage rolling toward states such as Virginia where landfill charges were cheaper. Tokyo, which in the 1990s generated three times as much solid waste as Mexico City, had virtually no health problems related to garbage. Indeed, the Japanese had enough money and ingenuity to find good uses for garbage, con- verting some into construction materials. (Note 33)

Considering all urban pollution and waste problems together, I hazard the following chronology and conclusions. Cities in the richer countries suffered serious sanitation problems as they grew rapidly. But by 1940 they had addressed these difficulties, through garbage collection, sewage

Note 32:
0n Surat, see WRI i997:42-3.

Note 33:
0n New York's landfill (25 times the size of the great pyramid at Giza in 1994), see Trefil 1994:23. Calcutta's garbage habits are described in Basu 1992. On American practices, see Melosi 1981.

P 289

systems, and water treatment plants. This left them with pollution derived from industry and transport, the scale of which continued to grow. After about 1970 (earlier in the case of smoke and soot), cities faced these prob- lems and often reduced their intensity. To some extent, rich cities "solved" their problems by shunting them off on downstream or downwind neigh- bors, but eventually this proved less practical as neighbors learned to use politics and the courts to prevent it. By and large, cities did not address pollution that threatened only diffuse, disorganized, or powerless com- munities. Cities in wealthy countries took about a century to organize partially effective responses to the pollution effects of urbanization.

Cities in poorer countries did not follow the same trajectory. In many cases they grew too fast, with too little accompanying economic growth, to afford clean water and garbage collection. The sanitary revolution that came to western Europe, Japan, and North America (c.1870-1920) came late and as yet incompletely to Brazil and South Africa, and scarcely at all to Bangladesh and Papua New Guinea, where urbanization rates were among the fastest in the world in the 1990s. And it scarcely came to the world's shantytowns, which spread exuberantly after 1950. (Note 34) Hence, in the latter half of the twentieth century, pollution derived from household wastes continued to grow in poor cities. In addition, poor cities rapidly ac- quired the pollution problems derived from industry and from fleets of cars, trucks, and buses. Djakarta in 1980 and Bombay in 1990 suffered a double burden of pollution almost never experienced in the rich world, except perhaps in the first industrial cities during the period 1820 to 1860. Only a few societies managed to accumulate capital for investment in pollution abatement at rates faster than pollution itself intensified. And even in those that did, ruling elites usually found it easier to insulate themselves from pollution rather than to reduce it. So cities remained concentrated nodes of pollution-far larger yet far less lethal than before, thanks to vaccination, antibiotics, and other public health measures. (Note 35)

THE ECOLOGICAL FOOTPRINTS OF CITIES. Cities themselves cov- ered perhaps 0.1 percent of the earths land surface in 1900 and about 1 percent in 1990. (Note 36) Their spatial growth, however, was only a small fraction

Note 34:
The growth and morphology of shantytowns (favelas) in Rio de Janeiro is traced in Abreu 1988:106, 125-6; see also Pineo and Baer 1998.

Note 35:
A survey of early 1990s conditions is given by Hardoy et al. 1993.

Note 36:
Douglas 1994 offers the figure used for 1990. I calculated the 1900 figure on the basis of a world population one-third as urban, one-fourth as large, and with a slightly higher density in ionn Hian in 1990.

P 290

of their environmental impact, for they spread their tentacles far and wide, drawing in food, water, and energy sometimes from other conti- nents. The space needed to support a city, and to absorb its wastes is, metaphorically speaking, its ecological footprint. (Note 37)

For millennia cities were biological oddities. They made good habitat for cats, rats, pigeons, and a small array of weeds, but they were shorn of most other animal and plant life. This changed little in the twentieth cen- tury. For each city that added a "green belt" as London did in 1936, or planned green spaces as Ankara did in the 19305, several lost what green space they had, as Mexico City did between 1950 and 1990. (Note 38) Perhaps the greatest biological change came with the gradual disappearance from cities of horses, camels, donkeys, and other beasts of burden. In physical as opposed to biological respects, cities often changed fun- damentally. In the wealthy world, cheaper and stronger steel allowed the construction of skyscrapers from early in the century. Electrification brought a myriad of changes, not least the electric trolley (invented in i887 in Richmond, Virginia), which encouraged suburbanization around the turn of the century. Then cars replaced horses and buses all but elim- inated trolleys. Electrification, sewage systems, and piped water con- verted the cities' substrate into a warren of tunnels, mains, and pipes. European and North American cities sprouted toward the sky, burrowed underground, and crept across the earths surface simultaneously. In 1870, most cities were held together by muscle and bone: people and horses carried or pulled all the food, water, goods, wastes, and informa- tion that circulated. By 1920, cities in the wealthy parts of the world (and a few elsewhere) were immensely complex systems of interlocking tech- nical systems. Not coincidentally, they were the powerhouses of the mod- ern economy and the hothouses of modernism in art and literature. (Note 39)

Such changes in cities' built environments came between 1880 and 1940 in the rich world. They came only a few years or decades later in colonial cities such as Bombay or Tunis. Many colonial cities acquired modern in- frastructure-ports, warehouses, railroads, roads-to meet the needs of

Note 37:
See Rees 1992.

Note 38:
The Wall Street Journal, 4 March 1993' reported that Mexico City had 21% tree cover in the 1950s and 2% in 1993- On Ankara, see Turkiye Cevre Sorunlan Vakfi 1991:50.

Note 39:
In 1939, New York City had 72,000 kilometers of underground mains, pipes, and ducts(Konvitz 1985:139). On environmental change and urban morphology in U.S. and European cities, see Ausubel and Herman 1988, Gugler 1996; Hurley 1997; Melosi 1990, 1993; Platt 1991; Relph 1987; St. Clair 1986; Vance 1990; and Whitehand 1987, 1992. On Latin American cities, see Pineo and Baer 1998.

p 291

an export economy. Some acquired sanitation infrastructure, usually only for parts of the city where Europeans or Japanese settled. British Singa- pore conformed to this pattern. The discrepancies between the Euro- pean and the Chinese or Malay parts of Singapore rankled, probably helping to motivate independent Singapore's singularly determined cam- paigns to achieve a thoroughly orderly and sanitary environment after 1960. Seoul, which fell under Japanese rule in 1910, appalled its new mas- ters with its filth and pollution, which they soon set about controlling. (Note 40)

After 1950, cities in all parts of the world increasingly spilled out over neighboring territory. In the United States, suburbanization became the dominant trend in urban life after 1945. By 1980, roughly two-thirds of the population in the 15 biggest metropolitan areas lived in far-flung, low-density, automobile-dependent suburbs. Cheap land, cheap cars, cheap gasoline, and mortgages made cheaper by the federal tax code contributed to the unusual American pattern. (Note 41) Elsewhere suburban sprawl more often featured government-financed apartment blocks or, more often still, shantytowns. Delhi covered 13 times as much land in 1990 as in 1900, engulfing a hundred villages and countless peasant plots in the process. Beijing doubled in extent in the 1990s alone. (Note 42)Suburban sprawl, whether in the form of genteel lawns and shopping malls or jer- rybuilt shacks, typically came at the expense of farmland. But while hun- dreds of millions of people took part in these socially momentous changes, the land directly subsumed came to only a few million hectares, equivalent roughly to the area of Costa Rica or West Virginia.

Urban growth nonetheless had tumultuous effects on water, land, and life-because cities have metabolisms. They take in water, food, oxygen (and more) and discard sewage, garbage, and carbon dioxide (and more). Fast-growing cities, like teenagers, have higher metabolisms than those that have stopped growing. Consider the daily diet of one leviathan. Hong Kong. In 1830 Hong Kong was an overgrown village. The collision be- tween China and Great Britain in the First Opium War (1839--1842) re- suited in over 150 years of British rule, during which Hong Kong prospered as one of the few conduits between China and the rest of the world. In 1900, Hong Kong had a quarter million people, in 1950 nearly

Note 40:
On Singapore, see Ho 1997 and Yeoh 1993; on Seoul, see Duus 1995. Other useful studies with reference to colonial cities are Abreu 1988 (on Rio de Janeiro), Coward 1988 (on Sydney), Kosambi 1986 (on Bombay), and Low and Yip 1984 (on Kuala Lumpur).

Note 41:
Jackson 1985.

Note 42:
Hardoy et al. 1933:115.

p 292

2 million. By 1971, when Hong Kong had 4 million people, its daily urban metabolism redirected prodigious flows (see Table 9.3). Seven-eighths of Hong Kongs food came from outside the Hong Kong Territory. A quar- ter of its fresh water was piped in from China, which in exchange took 40 tons of human excreta per day (for fertilizer). It required major adminis- trative and engineering feats to maintain Hong Kong-and greater ones in larger and more industrial cities like Beiiing, Tokyo, or Sao Paulo. (Note 43)

Hong Kong gets plenty of rain. Assuring other cities adequate water re- quired stronger measures, all the more so because of the physical char- acter of modern cities. Cities' roofs and roads prevented water from

Table 9.3 Daily metabolism of Hong Kong, 1971