5. The Initial Location of Africans and Europeans in Early America

The Voyages of Discovery and the interchange of biological organisms between the Old and New Worlds begin our in depth discussion of the roles that microbiology, evolutionary theory, and diseases have on our interpretation of the history and economic development of America.1 We explain the initial use of Europeans in colonial British North American agriculture, the eventual use and predominance of Africans (blacks) in the tropics (the Caribbean) and subtropics (the southern mainland) of colonial America, and the continued use and predominance of Europeans (whites) in the more temperate regions (the northern mainland) of colonial America.

The migration of Old World peoples—the term “Old World” refers to the Eurasian land mass, Africa, and the islands peripheral to these continents—to the New World set in motion an evolutionary process that led to very different regional disease ecologies in the Americas post-Contact. Our explanation for the predominance of blacks in the Caribbean and the southern mainland and whites in the northern mainland is in the existence of biological and epidemiological differences between populations in different regional disease environments. The explanation emphasizes the interactions among economics, changing populations, and disease environments. Diseases act on populations by causing sickness and death. But specific diseases do not afflict all ethnic (racial) groups equally; therein lies our story.2

The Consequences of Contact between the Old and New Worlds

Contact between the Old and New Worlds altered the ecological environments in both Eastern and Western Hemispheres. The hemispheres had been isolated from one another for geological ages except for the times during the Ice Ages when the Bering Land Bridge was in existence. The peoples who populated the New World in the sixteenth and seventeenth centuries were the descendants of the Ice Age migrants (American Aborigines or American Indians) and the more recent Old World migrants of Africans and Europeans, along with mixtures of these people. The demographic experience of these populations varied regionally.

Native Americans

The first human inhabitants of the New World are generally believed to have migrated from northeastern Asia (Siberia) to the North American continent. Wallace (cited in Bishop 1993, p. 1), using evidence derived from DNA analysis, argues that humans also migrated from Polynesia to the Americas, but little is known about these migrants. The Siberian migrants crossed over on the Bering Land Bridge that connected northeastern Asia to the Alaskan peninsula, or along the near-shore coastal waters to its south, during the last Ice Age glacial about 15,000 or more years ago.3 Humans migrated south until finally Aboriginal Americans inhabited essentially all of North and South America. The only major uninhabited exception was Barbados, the most eastern island in the West Indies.

They came relatively free of disease. The absence of disease was a result of the frigid climate, the small absolute size of the migrant bands, and the absence of domestic animals except the dog. Hunter-gatherer groups usually number from 25 to 75 individuals; this means that for most infectious diseases the absolute number of human beings in any tribal group was not large enough to enable infectious diseases to survive. The lack of domesticated animals prevented the Amerindians from contracting most zoonotic diseases. (Measles, influenza, and chicken pox are examples of diseases that have crossed over from animal populations to infect humans.) Ice Age temperature ensured that airborne pathogens quickly died if they had no host. (In hot and humid climates, the tropics, some airborne pathogens can survive for days without a host.) As a result of these factors, prehistoric humanity was relatively disease free compared to civilized humanity—living in dense, permanent clusters (McNeill 1976).

After the end of the last Ice Age glacial and the subsequent disappearance of the Land Bridge, Aboriginal Americans were almost completely isolated from contact with the Old World and remained relatively (compared to the Old World) disease free for the millennia prior to Contact. Isolation had two significant effects: (1) New World populations exhibited much less genetic diversity than Old World populations, and (2) the lack of domestic animals and communications with other peoples meant that the New World had a disease environment much less complex than the Old World. For millennia, Old World peoples had exchanged diseases and contracted diseases from domesticated animals and wildlife. This process had almost no counterpart in the New World. When animals were domesticated in the Old World, zoonoses (chicken pox, measles, probably smallpox, and yellow fever, among others) were passed from animal populations to human populations. Over centuries, they were transformed from epidemic diseases that affected all age groups to endemic or so-called childhood diseases in Old World human populations.4 The isolation of the New World meant that when these diseases were introduced into the New World they were epidemic (“killer”) diseases. Many other diseases were transmitted from Old World peoples to the New World; there were few major diseases (maybe only one, syphilis) that were endemic in the New World that were unknown in the Old (Diamond 1997; McNeill 1976). Consequently, when interactions on a regular basis with the Eastern Hemisphere began, there was an almost immediate population implosion of Aboriginal Americans due to “virgin soil” epidemics. Because of the relatively benign disease ecologies of the New World, there was no corresponding population implosion in the Old World.

We emphasize the word “relatively.” The New World was not disease free (Buikstra 1993; Karasch 1993). Indeed, there are substantial disputes over the possible American origins of various pathogens; the best known of these is syphilis. In a recent study, Harper, Ocampo, Steiner, et al. (2008) indicate that New World origins of syphilis, rather than Eurasia or Africa, are consistent with DNA examinations of bacteria in the syphilis/yaws family. Regardless of its origins, in the Old World syphilis was a rapidly fatal disease for many of its victims when first introduced into late fifteenth-century (circa 1495) Europe; over time both the pathogen and the populations adapted to one another.

The Voyages of Discovery brought diseases to the New World to which Aboriginal Americans had no previous exposure. Columbus’s second voyage to the New World in 1493 started an epidemic that began the extinction of the Caribbean Aborigines. Guerra (1988) contends that this epidemic was influenza contracted from the swine Columbus acquired when he stopped for provisions in the Canary Islands. Regardless what it was, it killed: Guerra (1988, p. 319) estimates the aboriginal population of Hispaniola declined from over one million to just 10,000 from 1492 to 1517. These estimates imply that the mean annual decline in the Amerindian population from 1492 to 1517 was a little over 18 percent per year. Dobyns (1983) dates the first New World smallpox epidemic at about 1516. Consequently, most of the decline in the aboriginal population of Hispaniola took place before the introduction of smallpox.

Native peoples of the Caribbean, Mesoamerican, and Andean civilizations were devastated by exposure to Old World diseases.5 The tropical pathogens imported from the Old World (along with measles, smallpox, and influenza) were especially devastating to the inhabitants of the tropical and semitropical regions of the Americas. Brooks (1993) offers a revisionist version of the impact of smallpox, contending that smallpox was neither as contagious nor deadly as other scholars would have us believe. Brooks has been severely criticized by McCaa (1995), who uses the sources that Brooks relied on to undermine Brooks's hypothesis. Regardless of these differences, over the period 1518 to 1568, the data favored by Brooks give an annual decrease in population in Mesoamerica of over 2 percent; over two centuries this would produce a population decline of considerably greater than 95 percent! As a result, except for isolated areas in the Amazon and Orinoco River basins, no pure Aboriginal Americans exist in the tropics of the New World.

Recently, the pre-Contact population estimates of Aboriginal Americans have been subject to a continuous stream of upward revisions. In particular, the pre-Contact population of the Amazon aboriginals has been revised upward substantially; some partisans in the dispute even have suggested that parts of the Amazon were as densely populated as the fifteenth century’s Old World rice civilizations.6

Many scholars believe that for most practical purposes, pre-Contact Americas were virtually a virgin soil for Old World diseases.7 Others contend that the New World was more exposed to diseases than the “virgin soil” scholars portray; the recent DNA analysis of syphilis that points to its New World origins lends some credence to this school. Still syphilis is the only New World disease that was a major infection for Old World peoples; conversely, there were many Old World diseases that had significant effects and were new to the indigenous Americans. The catalog of pathogens brought from the Old to the New World is lengthy, but not complete; discoveries are still being made. A nonexhaustive list includes smallpox, influenza, measles, malaria, yellow fever, pneumonia, intestinal parasites, ectoparasites, dengue fever, onchocerciasis, trachoma, and leprosy; not all of these diseases are unanimously accepted. Cockburn (1980, p. 159) believes that hookworm was definitely resident in the New World prior to 1492 because of microscopic evidence found in Andean mummies. Hurtado, Hurtado, and Hill (2004, p. 173) dispute this, arguing that contemporary indigenous American peoples “are hyperinfested with helminthes and ectoparasites”; carrying a heavy disease load relative to others is not consistent with them having a long evolutionary history of exposure to these parasites.

The epidemics that devastated the New World also affected the relative mix of Native Americans. Those who had developed relatively densely populated societies based primarily on agriculture were adversely affected relative to the societies that were less developed and more nomadic (hunter-gatherers). Prior to Contact, the nomadic tribes were marginal and inconsequential relative to the total aboriginal population. The densely populated societies of the New World facilitated the spread of Old World pathogens that effectively destroyed the relatively developed aboriginal societies. Societies that followed a more nomadic lifestyle and used natural resources less intensively were less severely affected. Fewer people meant less hunting, and less hunting meant that populations of fish and game exploded. Given the short (relative to humans) generational time in game animal populations, the increase in the wildlife available to surviving Native Americans subsequent to Contact was substantial. Some Old World animals (cattle, horses, sheep, and swine) that were imported became feral and endemic to the American landscape, providing additional food resources. Skeletal remains show that Native Americans prior to Contact were very short and suffered from bone deformities associated with overwork and/or nutritional deficiencies. Post-Contact skeletal remains, although much fewer in number, show that Native Americans had fewer deformities and were substantially taller (Larsen, Crosby, Griffin, et al. 2002).

Figure 5.1 illustrates our conception of the effects that Contact with the Old World had on the native populations of the New World. Contact devastated the Aboriginal American populations. The arrows in figure 5.1 indicate the direction of causality and the algebraic signs indicate the historical effect that occurred as a consequence of Contact; the size of the arrows proxies what we estimate to be the magnitude of the effects. Starting from the top, Contact reduced the total population of Native Americans, and the large decline in population had sizeable adverse effects on densely populated societies (as indicated by the larger black arrows connecting the boxes) relative to less densely populated societies. Contact also increased the demand for labor on the part of the conquerors, which led to the enslavement of the Native Americans. In turn, enslavement had a sizable adverse effect on densely populated societies relative to the more nomadic hunter-gatherer societies.8 Nomadic societies became more attractive to Native Americans who had lived in relatively sedentary agricultural communities because of the increase in animal populations; this is shown in figure 5.1 by the negative effect that an increase in animal and fish populations had on the “Densely populated native societies.” To emphasize a point, besides the increase in animals native to North America, feral Old World species augmented the supplies of animal proteins available. Depending on hunting for a major portion of one’s food is typically a very attractive option to relatively primitive agriculturalists. Europeans also were attracted to the lifestyle of hunting civilizations. There are numerous accounts of Europeans joining native societies because they found the lifestyle attractive. People became agriculturalists in lieu of hunting only when game supplies were reduced.9

The demographic experience of Native Americans in the mainland North American colonies was similar; populations decreased rapidly after British settlement (this was in addition to the earlier massive die off that occurred almost immediately post-Contact). Smallpox was the ubiquitous and frequently deadly common denominator to all disease environments in the post-Contact New World. However, with the significant exception of smallpox, the Old World pathogens that afflicted the natives on the British colonial mainland were not identical to those in the Caribbean. In addition, the disease ecology in the mainland colonies varied as the climate varied. In New England, cold-weather diseases predominated; diseases such as influenza, tuberculosis, pleurisy, pneumonia, and other lung infections. The disease environment changed from that of primarily cold-weather diseases to a mixture of cold-weather and warm-weather diseases as one traveled south from New England on the mainland. In the colonies of Virginia, Maryland, and North and South Carolina, there were no diseases that were distinctly “their” diseases; instead the disease environment was more like an amalgam of the diseases that were in New England and in the tropics. These diseases were neither as virulent as in their native climates, nor were the human inhabitants as severely afflicted by any one disease; still there were more diseases to cause sickness and death. In the tropical regions of British America, the disease environment consisted almost entirely of warm-weather diseases: malaria, yellow fever, dengue fever, hookworm, schistosomiasis, and the ubiquitous smallpox.

Smallpox was an exclusively human disease; there were no nonhuman reservoirs or vectors for smallpox. Consequently, when smallpox was introduced into a region or people, it was always through human intervention. This explains smallpox's lack of regional or geographic specificity. Smallpox was a truly human disease; it went wherever people migrated (Crosby 1993b).

Given enough time the population of Native Americans could have recovered from the introduction of Old World pathogens to become, once again, the dominant ethnic group of the Americas. But the Aboriginal American population was not given sufficient time: diseases, enslavement, and the attractions of hunting-gathering lifestyles combined to ensure that Native American civilizations did not have time enough to recover. In the three plus centuries of the colonial period, growing populations of Africans and Europeans and their descendants displaced Native Americans as dominant ethnic groups in the New World.

Genetic Homogeneity: The Chromosomal Evidence

Microbiology, DNA analysis, and evolutionary theory offer explanations for the devastating impact Old World diseases had on the New World population. The peoples of the New World were, relative to Old World populations, genetically homogeneous. There are historical reasons for the lack of genetic diversity in the peoples of the New World (Wang, Lewis, Jakobsson, et al. 2007). After migrating from Siberia during the last Ice Age glacial, Aboriginal Americans were isolated for millennia from other humans. Their isolation from the Old World led to a genetically homogeneous population; their lack of genetic diversity is supported by evidence from an examination of immune systems. The locus (chromosomal position) of a part of the immune system, the major histocompatibility complex antigens (termed Class I and Class II MHC glycoproteins) has been identified. The chromosomal position establishes that Class I and Class II MHC glycoproteins are transmitted genetically. An individual's immune system relies on these glycoproteins to identify and assist in the destruction of pathogens. (There is a Class III MHC glycoprotein that encodes for other immune components; since it is not critical to our analysis, it is not discussed here.)

How the body defends itself depends on the nature of the pathogen. Large bacteria and helminthic (parasitic worm) infections are present outside the cell (extra cellular). These pathogens are directly attacked by the immune system. The system deploys antibodies (soluble antigen receptors) that bind directly to the pathogen and act as a signal for other parts of the immune system to aid in the pathogen's destruction. Smaller bacteria, protozoa, ricksettia, and viruses establish their infections inside the cells of humans. To attack these intercellular pathogens is the function of MHC glycoproteins. All cells carry MHC molecules on themselves; on a cell infected by a pathogen, the MHC molecules display a fragment of the pathogen that acts as a signal for further action by the body's immune system. How the immune system wards off invaders depends on a subset of the MHC glycoproteins, which are genes, and are referred to as human leukocyte antigens (HLA); these are genes that attack various bodily invaders. There are large numbers of HLA alleles (alternative forms of a gene) that can concentrate on specific types of invaders. Although Class I and Class II MHC glycoproteins (and their genetic subcomponents, HLA) are not completely understood, it appears that Class I MHC glycoproteins attack invading microorganisms while Class II MHC glycoproteins control damaged or mutant cells within the body (Black 1992, 2004; Englehard 1994).

Class I and Class II MHC glycoproteins are recognized as the body's defenders against cellular invasion. Evolutionary selection favors viral (and other pathogenic) infections that survive the immune system's counter attack. Pathogens are subject to culling by the body’s immune system; pathogens that survive and reproduce in the body have been relatively effective in subverting attacks by the host’s immune system. A particular strain of a pathogen can be relatively successful in combating the onslaughts of a specific individual who has a limited set of Class I and Class II MHC glycoproteins. If sick people infect genetically related individuals with pathogens that survived attacks by their human immune system, then the pathogens are pre-adapted to the defense mechanisms of their new hosts. This is because both the new victim and the source victim have very similar Class I and Class II MHC glycoproteins as a result of their common genetic heritage.

An individual can have up to six different kinds of Class I MHC glycoproteins; populations, in general, have many more types of Class I MHC glycoproteins than an individual. Engelhard (1994, p. 56) states that over 100 forms of MHC glycoproteins have been identified; the number of HLA alleles is vast and continually being upgraded. Here we focus on MHC glycoproteins, while recognizing that the HLA are the proximate defenders of the human body. In recent studies, 40 different types of Class I MHC glycoproteins have been identified in a population of sub-Saharan Africans, 37 in a population of Europeans, 34 in a population of East Asians, but only 17 in a population of North American Indians, and 10 in a population of South American Indians (Black 1992, 2004). The lack of genetic diversity among North and South American Aborigines meant that a pathogen could spread more rapidly through New World populations than through Old World populations. This explains the devastating impact of Old World diseases.

Consequences of Genetic Homogeneity

Black (1992, p. 1739) reports that as recently as the 1970s isolated people in the Amazon experienced mortality rates of as high as 75 percent when exposed to relatively common diseases—measles, mumps, and influenza. The isolation of the New World aborigines that kept them free of Old World pathogens also led to their devastation for a related reason. It meant that their ancestors were not selected for their resistance to smallpox, measles, mumps, yellow fever, and influenza, among other diseases. Ramenofsky (1993, p. 322) estimates that approximately 14 diseases were introduced into the Americas between 1500 and 1700.

The argument is one of population genetics, not individual genetics: two individuals from different populations may face the same risk (probability of death), but the population that has the greater genetic diversity within it will be at less risk. With a South American Indian population that has 10 Class I MHC glycoproteins, the chances are 32 percent that a virus passing between two people will not encounter a new type of Class I MHC glycoprotein. Among sub-Saharan Africans with 40 different Class I MHC glycoproteins, the chances of a virus not encountering a different Class I MHC is 0.5 percent (Black 1992, p. 1739). This means that pathogens are much more virulent among peoples with little Class I MHC diversity (American Aborigines) than among populations with a relatively large number of different Class I MHC glycolproteins (Africans and Europeans). Hurtado and Salzano (2004, table 11.1, pp. 213–15) show a case mortality rate (percentage of individuals with a disease that die) of 17.7 percent for a measles epidemic among the Yanomamo people (South American Amazon natives) in the 1960s.

An additional problem is that many diseases that kill adults are relatively mild when afflicting children. Yellow fever, measles, chicken pox, and mumps are examples of diseases that mildly impact children but are mortal threats when contracted as an adult (Cooper and Kiple 1993; Kim-Farley 1993). American Indians faced three reinforcing disasters because of epidemics: The obvious first was that many adults died; the second was that with the adult deaths, children, who would have survived with care, perished without adult care; and the third was the death of so many prime-aged adults reduced the resources available to the populations and left them vulnerable to starvation, conquest, and assimilation. Adults similarly died from lack of nursing—that is, from dehydration and disease brought on by the lack of sanitation (Dollar 1977).10

Although fewer data exist on morbidity, in general, we expect that the indigenous American populations were very sickly. It is likely there was a positive correlation between morbidity and mortality rates for the types of Old World diseases to which the natives were susceptible. This would reduce their effectiveness in supplying agricultural labor. As a result, the increased morbidity and mortality of American Indians caused by Old World diseases decreased their lifetime productivity, and consequently decreased the demand for them as unskilled agricultural workers. The negative implications for natives after Contact was so deleterious that the demand for Native Americans as bound servile agricultural labor became negligible in the mainland British North American colonies. The existence of a slave trade in American Indians in the Carolinas in the late seventeenth and early eighteenth centuries proves the point. It failed to survive because Native American populations were declining and the demand for them was decreasing because of their excessive morbidity and mortality (Gallay 2002).

The decline in New World populations presented the Old World conquerors with a dilemma: How were they to get labor to exploit the resources available in the New World? The thought of doing the work themselves was not palatable to the conquerors. A variety of solutions to the problem of labor supply in the New World were attempted. The Conquistadors and subsequent invaders first tried the immediate expedient of enslaving distant Aboriginal Americans after the nearby ones died; after these Aboriginal Americans also died (similar to the local natives, they typically died of Old World pathogens), other sources were used. Various ethnic and religious groups were sent as servile labor from the Old World. “Servile” includes indentured servants who negotiated their contracts and other laborers who were not free to negotiate their own contracts. Some of the latter were slaves, prisoners of war or conscience, and convicts. The result of these “experiments” was that by the middle of the eighteenth century populations of differing racial (ethnic) backgrounds were geographically divided. In general, in the temperate regions of the conquered Americas, Europeans predominated; Africans predominated in the tropical regions. Only in the high mountain regions and high plains (and in the unconquered areas) did Native Americans predominate. In the late twentieth century, Latin American (South America and Mesoamerica) peoples that were identified as indigenous accounted for 3.3 percent of the total Latin American population; only Peru (40.8 percent), Ecuador (43.8 percent), and Bolivia (56.8 percent) had an indigenous population of more than 15 percent of the total population (Hurtado, Hurtado, and Hill 2004, p. 167).

These are generalities, the specifics of demographic change depended on geography and history along with biology. In the United States, the land east of the Mississippi is usually considered a temperate region, but there are certainly significant variations in climate within this area. In this region, the more one travels from north to south, the more the climate variables make the area a “warm-weather” region. In this context, we conceive of climates and regions as a continuum between “cold weather” and “warm weather”: the more frost free days there are, the more likely the climate of a region approaches the “warm weather” category or classification. A more nuanced statement concerning geographical divisions and ethnic groups is that Europeans predominated in “cold-weather” regions and Africans predominated in “warm-weather” regions of the conquered Americas. The experiences of Old World populations in the New World disease ecologies differed radically from the disasters suffered by the indigenous American populations.

Africans and Europeans

The Caribbean and Tropical Africa

The demographic experience of Africans and Europeans in the tropics of the British New World (the Caribbean), and in tropical Africa, differed from their experience on the British colonial mainland. Involvement in tropical Africa brought Europeans into contact with pathogens for which they had little or no prior exposure. The African pathogens that most afflicted Europeans were hookworm, malaria, and yellow fever, which along with other tropical pathogens, combined to produce phenomenal rates of mortality and morbidity for Europeans in tropical Africa. Examples of other tropical diseases that adversely affected Europeans relative to Africans and thus affected the choice of labor supply were yaws, dengue fever, and schistosomiasis. If we were to include the impact of all of these diseases, it would reinforce our contention about the relative impact of tropical diseases. After tropical pathogens were established in the Caribbean, death rates for non–West Africans in the Caribbean were a multiple of the rates of people of African origins until well after the mid-nineteenth century. Tropical diseases served to eliminate substantial numbers of Europeans from the sugar-producing regions of the New World tropics, resulting in present-day populations in the former Caribbean colonies of Britain, France, and Holland that are primarily of African origins. The former colonies of Spain (Cuba, Puerto Rico, and Santo Domingo) are exceptions. These islands have retained a substantial European contingent in their populations. There are various explanations for this, among the explanations are an Iberian resistance to African diseases; the settlement of cities, which allowed people of European ancestry to avoid malaria and become resistant to yellow fever; and the laggard development of full-scale sugar planting on these islands.

Hookworm, malaria, yellow fever, and other Old World tropical diseases were introduced to the Caribbean after the Voyages of Discovery. Hookworm of the Necator americanus variety (one of the two primary hookworms that afflict humans, the other is Ancylostoma duodenale) was almost certainly introduced when Africans were first brought to the New World (in the late fifteenth and early sixteenth centuries), but its effects are hard to trace. This is because hookworm is a debilitating disease but not frequently deadly. An individual with a severe case is likely to die of something else. And the etiology of hookworm in the New World was unknown and unstudied until the early twentieth century (Ettling 1981). Similarly, malaria was introduced almost certainly shortly after the first Voyages of Discovery in the late fifteenth century, but because its symptoms can be confused with other fevers, it is hard to date its introduction exactly (Kiple 1984, 1993).

While specific dating of the introduction of hookworm and malaria is not possible, the first yellow fever epidemic affecting Europeans in the Caribbean has been studied and is dated at 1647 in Barbados (Findlay 1941, p. 144). Prior to the mid-seventeenth century, Barbados and the Caribbean in general were not regarded as particularly unhealthy for Europeans. Henry Colt, a British “tourist” in the West Indies in the early seventeenth century, was struck by the absence of disease (Findlay 1941, p. 151). Evidence from other Europeans during the first 150 years after the Voyages of Discovery suggests that the disease environment of the Caribbean was unremarkable for Europeans. Prior to 1647, yellow fever was unknown; after 1647, it became endemic. The permanent establishment of tropical West African pathogens, and the subsequent increase in European mortality rates circa 1650, are consistent with the employment of primarily European servants until the late 1640s and the switch to primarily African slaves during the 1650s in Barbados, and the Caribbean in general. Kiple and Higgins (1992) suggest that yellow fever was what changed the sugar-producing Caribbean's population from European to African (for a history of Caribbean sugar plantations, see Dunn 1972).

The importance of yellow fever in affecting labor choices outside the sugar-producing Caribbean is debatable. First, it is a disease whose vector, the mosquito Aedes aegypti, has a limited flight range estimated to be a maximum of a few hundred yards and is considered a disease that thrives in centers of dense human populations (sugar plantations or urban areas). Second, the case mortality rate is debatable (estimates range from 20 to 70 percent with a much lower mortality rate for children), and a victim of yellow fever who recovers enjoys lifetime immunity (Cooper and Kiple 1993, p. 1100). Third, the generations of Europeans growing up in the Caribbean would have acquired immunity to the disease by childhood exposure, as did the Spaniards of Cuba and Puerto Rico. Thus, yellow fever alone does not adequately explain the transition from a European to an African agricultural labor force in the non–sugar-producing Caribbean.11

Although neither hookworm nor malaria is as deadly as yellow fever, both are debilitating, and their study is instructive. (We postpone the discussion of hookworm until we discuss the Chesapeake region because hookworm’s discovery and history are inextricably linked with the American South.) Four varieties of malaria have been discovered; one, falciparum malaria (Plasmodium falciparum), is frequently deadly. The other types are less deadly (but still debilitating) and more geographically diverse than P. falciparum. Exposure to one variety of malaria gives no immunity to the other varieties, but attacks of other varieties may be less virulent in people who have suffered from previous malarial attacks. We concentrate on P. falciparum because it is the variety of malaria that is most frequently deadly. It is death that is most effective in increasing the relative frequencies of HLA alleles, which resist a disease in a population. The mechanism is that people with resistance are more likely to survive long enough to reproduce, while those who lack resistance are more likely to die before reproducing. Thus, killing diseases are more effective in altering the genetic composition of a population than are high-morbidity diseases. P. falciparum is endemic to the tropics of West Africa, and within the human populations in Africa innate resistance has evolved. A substantial portion of the tropical West African population has the sickle cell trait and/or other blood abnormalities that provide protection against malarial attacks. These traits are uncommon in populations not exposed to P. falciparum but do occur in both African and non-African populations where malaria and P. falciparum are endemic (Kiple and King 1981, pp. 17–22).

The reasons for the extraordinarily high susceptibility of Europeans to tropical diseases and African resistance to them lie in microbiology and evolutionary theory: in an environment with endemic hookworm, malaria, and yellow fever innate resistance is pro-adaptive. A gene that confers resistance to these pathogens will allow the humans who possess it to lead longer and healthier lives, other factors the same. The “other factors the same” condition usually does not hold when speaking about the impact of genes. Genes typically have more than one effect on the body; these phenotypic effects have long been noted. Darwin called them “correlations” and wrote that they can be “quite whimsical: thus cats with blue eyes are invariably deaf; . . . [h]airless dogs have imperfect teeth; . . . pigeons with feathered feet have skin between their outer toes” (as quoted in Cronin 1991, p. 60). Some phenotypic changes will have positive survival value, others negative, and some may have no effect. A statement that is more consistent with the typical effects of genes would be that a gene that confers resistance to the prevailing disease organisms in a given environment, and whose other phenotypic effects do not completely offset their survival advantage, will spread in the relevant population. Each generation that carries this gene will, in all probability, leave more descendants than their peers who do not possess the gene. Over generations and millennia, the gene will tend to become more frequent in the population. The epidemiological results of this process are dramatic.

Acquired childhood immunities also play a part. In an environment where a disease is endemic, the immune system will have had early childhood exposure to the disease-bearing microorganisms. As a result, the immune system will be able to quickly produce antibodies to any subsequent exposure.

Figure 5.2 illustrates the impact of the Old World migrations on the disease ecology in the warmer regions of the New World. As before, the arrows in the schematic diagram indicate the direction of causality. Starting from the top, Europeans’ demand for agricultural labor in the British New World leads to migrations of Europeans and importations of African slaves to America. Africans and the ships that transported them carry warm-weather diseases (hookworm, malaria, yellow fever, and other tropical pathogens) with them. With chance and sufficient time, these diseases become endemic. The warm-weather diseases adversely impact European agricultural laborers who become sickly and possibly die. Imported African slaves were very likely to have had previous exposure to the warm-weather diseases imported from the tropics of West Africa and consequently would have had acquired immunities; they also were very likely to have inherited some innate immunities from their West African parents. The relatively greater illnesses and deaths among the Europeans would reduce their productivity relative to that of African slaves and, as a result, reduce the demand for Europeans as agricultural field hands. This then increases the demand for Africans as agricultural field hands, increasing slave importations. This feedback restarts the cycle, leading to a decline in the number of Europeans and their descendants living in the tropics of the British New World (the Caribbean).

Data collected by Curtin (1968) allows a comparison of the death rates of troops of differing ethnic origins in the British Army. Curtin’s data, which are summarized in table 5.1, are deaths rates for European troops (those enlisted in Europe and presumably primarily from the United Kingdom) and African troops (those enlisted in Africa or from captured slave ships leaving Africa). While the data are for the early 1800s, they are nevertheless relevant for earlier times. Since only fragmentary data exist for earlier periods, the assertion that Curtin's data are apropos demands a foundation. Recall the earlier discussion of human immune systems. The MHC glycoproteins that individuals possess are inherited from their parents. In a closed population succeeding generations will have the same number of different glycoproteins in their immune systems as the ancestor population had, unless only a small and biased sample of the ancestor population produced offspring. We know that the European population from the seventeenth through the nineteenth centuries was not closed; it incorporated people from all over the world into it; Asians, Africans, and American Indians did interbreed with elements of the European population. Additionally, there is absolutely no reason to believe that only a small and biased sample of the population reproduced. This means that, if anything, the European population of nineteenth-century Europe had a greater variety of MHC glycoproteins than its ancestor populations of the seventeenth and eighteenth centuries. If we consider only the British population as the relevant population, the argument is much stronger. From 1650 to 1850, the British population incorporated into it people from all over the world and people from other areas of Europe. Consequently, it would be extraordinarily unlikely that the British population in 1850 had a smaller variety of MHC glycoproteins than the British population of 1650.

^{Table 5.1: Death rates per 1,000 troops of different origins in the British army. Source: Curtin (1968, tables 1–3)}

The result is that, immunologically speaking, the nineteenth-century European population had the same, or greater, resistance to diseases than its seventeenth-century ancestor population. Consequently, Curtin's nineteenth century data are, if anything, a lower bound for the European experience in tropical environments in earlier times. This implication would have to be modified if the nature of medical care deteriorated over time or the disease environment became more deleterious. There is no evidence for the former, and this book argues against the latter.

Table 5.1 suggests that Africa was close to a death sentence for Europeans; the Caribbean, while better than Africa for Europeans, was decidedly unhealthful. African troops in tropical areas did better than European troops. (Repeating and stressing a point, these were African troops in the British Army.) Curtin (1968, p. 206) also includes mortality data for Jamaican born slaves (over three years old) of African ancestry for the years 1803 to 1817. He reports a death rate of 25 per thousand, a lower death rate than the lowest death rate for African troops serving in the West Indies. Perhaps surprisingly, African troops in the New World had mortality rates that were, on average, only approximately 25 percent higher than their mortality in Africa. The death rate for European troops in the Caribbean was at least eight times higher than the estimated rate of 15.3 per thousand in the United Kingdom (Curtin 1968, p. 202). The differential between African and European mortality rates is a result of the disease environment that prevailed in the tropics of the New World after contact with Africa was established on a regular basis. The disease environment was closely related to that of tropical Africa, to which Africans had been pre-adapted by natural selection and childhood immunities.

Curtin's data are based on British Parliamentary records and are acknowledged to be the best currently available. There are other estimates based on good data that, while also for years subsequent to the founding of the first colonies, should be considered lower bounds as well. Steckel and Jensen (1986) present reliable estimates of death rates for slave (African) and crew (European) during loading of the enslaved in Africa and during the voyages; these data are for the late eighteenth century. Steckel and Jensen's (1986, pp. 60–62) reported death rate for Africans while loading the slave cargo is 45.3 per thousand, and during the voyage 115.7 per thousand; the crew (primarily of European ancestry) death rates are 237.9 during loading and 207.4 during the passage. Davies (1975) provides another source of reliable data for the 1665 to 1722 period. Davies’s (1975, pp. 93–95) data come from the records of the Royal African Company; the data show first-year death rates for Europeans newly arrived in West Africa. The mean annual first-year death rate for these data is 62.8 percent (628 per thousand); the range in observation is from a low of 54.0 percent to a high of 66.7 percent. Davies (1975, p. 93) also notes that of the new European arrivals during the period: “one man in three died in the first four months in Africa, more than three men in five in the first year.”

Less precise and more impressionistic estimates for the death rates of Europeans in the tropics are available from a variety of other sources; they suggest that, if anything, the death rates for European troops in the Caribbean that Curtin presents—roughly from 120 to 140 deaths per thousand troops per year—are too low. Rogoziński (1992, p. 167) estimates that of the roughly 20,000 British troops that invaded Saint-Dominique (Haiti) beginning in September 1793 and continuing until 1798: “Almost 13,000 men died—perhaps 1,000 in battle, the rest of malaria and yellow fever.” Twelve thousand deaths from disease over a five-year period imply a death rate of approximately 200 per thousand per year.12 Writing about the later French invasion of Saint- Dominique, Rogoziński (1992, p. 172) states that of the 20,000 French troops that had landed by February of 1802: “By the middle of April half of the soldiers were dead or sick of yellow fever.”

On the African side during the eighteenth century, Miller (1988, p. 286) writes that: “They [men of European birth] settled in the commercial quarter of the lower city [Luanda in Angola]; then within weeks or months the majority was carried off to the European graveyard. . . . The pervasiveness of death . . . in Luanda . . . set the tumultuous style of business in a city that truly merited the sobriquet of the ‘white man's grave.’” The earlier experiences of Europeans with Africa were just as deadly; in the sixteenth century, the papal ambassador to Portugal accused the Portuguese Crown of violating the ban on the execution of Catholic prelates by simply exiling the offending churchmen to Africa “knowing that within a short time they would be dead” (Thornton 1992, p. 142).

McDaniel (1994) has argued that it was childhood-acquired immunities, not any innate resistance, that made West Africans more resistant to the diseases of tropical West Africa. Employing data from the migration of African-Americans to Liberia during the nineteenth century, McDaniel (1994, p. 106) reports crude death rates ranging from a high of 143 per thousand during the 1820 to 1828 period to a low of 70.7 per thousand during the 1836 to 1843 period. McDaniel uses these data to argue that innate resistance to diseases was not important, and that tropical West Africans do not have any inherent biological resistances to these diseases. The data that McDaniel presents indicate that the highest death rate for African-Americans in Liberia (143 per thousand) was less than one-third of the lowest death rate for Europeans in Africa (483 per thousand) reported in table 5.1. Except for the initial experience during the years 1820 to 1828, the death rate for African-Americans in Liberia was substantially less than for European troops in the West Indies during the same period (see table 5.1). McDaniel’s (1994, p. 106) mortality data for African-Americans in Liberia for the entire 1820 to 1843 period average 83 deaths per thousand; the simple mean death rate for European troops in the West Indies reported in table 5.1 is 132 per thousand. These data indicate that African-Americans in Liberia had 49 (37 percent) fewer deaths per thousand than did European troops in the British West Indies. Also note from table 5.1 that the simple mean death rate for African troops in the West Indies is 40 per thousand, less than one-third that of European troops (132 per thousand). The data indicate that while acquired antibodies may have played a part in the resistance of Africans in Africa to the local disease ecology, the relative resilience of African-Americans (born and raised in the New World) to the disease ecology of nineteenth-century Liberia was genetically transmitted.

McDaniel's (1994) case is further undermined by his own observation elsewhere that African-Americans of mixed ancestry (African and European) did much worse in Liberia than those of unmixed African ancestry (McDaniel 1995). “Contemporary observers suggested that mulattos were the leaders [of Liberian society] but believed that they suffered from excess mortality” (McDaniel 1995, p. 58). McDaniel (1995, p. 58) also observes that the migrants to Liberia “may have received assistance from their former masters because of past service.” An explanation for these payments other than “past service” would be that they were blood relatives: the children and grandchildren of slaveowners.

While McDaniel's (1994, 1995) evidence supports our contention of innate resistance and undermines his contention of none, the data are not entirely adequate to address the acquired immunity/innate resistance issue. Many of the migrants were of mixed ancestry, and death rates were not identified by ancestry. Indeed, given the absence of DNA testing, it is impossible to settle the issue once and for all because the only classification available in the nineteenth century would have been by appearance. Yet the visible physical characteristics of individuals may be entirely unrelated to their inherited immune system; thus, an individual who appears “African” may have a set of MHC glycoproteins that are more closely “European” than “African.”13

Quite obviously, the death rates of Africans in Africa were substantially lower than those of Europeans in Africa; otherwise, there would have been no Africans. The African slave trade opened the tropics of the New World to African pathogens. These diseases (dengue fever, hookworm, malaria, and yellow fever are prominent examples) afflicted Europeans more severely than Africans. Diseases and an African labor force reinforced themselves: the African slave trade brought into the New World tropical African pathogens and made European agricultural labor a less viable option in regions where these pathogens flourished.14

The British Colonial Mainland: The North

The demographic experiences of Africans, Europeans, and their descendants in the British mainland colonies were not only different from that in the Caribbean but also differed across areas of the mainland because the combination of diseases on the mainland varied regionally, and the diseases had differential effects on Africans, Europeans, and their descendants. Klepp (1994) provides evidence on differences in mortality of peoples of different ancestral origins during the eighteenth century. While Klepp focuses on eighteenth-century Philadelphia, her study sheds light on the differential impact that diseases had on different peoples and allows us to explain the predominance of Europeans in the northern colonies. Table 5.2 summarizes Klepp’s (1994) evidence as well as the mortality rates presented in Warren (1997). Combining Klepp's crude death rate data for cold-weather diseases (measles, pleurisy/influenza, and whooping cough) and smallpox in Philadelphia, the simple mean crude death rate during epidemic years for Africans and African-Americans (blacks) was 87 percent higher than the death rate during epidemic years for Europeans and European-Americans (whites) (see table 5.2, row 5). During epidemics of non–cold-weather diseases (yellow fever, diphtheria/croup, dengue fever, scarlet fever/scarlatina, and typhus/typhoid fever), the black mortality rate was only 10 percent higher than that of whites (see table 5.2, row 6).

Klepp (1994, p. 479) argues that black mortality reached a seasonal peak in the winter months, whereas white mortality reached a seasonal peak during the summer months. Her data imply that although white mortality was substantially lower than black mortality for the entire year in Philadelphia (see table 5.2, rows 7, 8, and 10), white mortality was considerably higher than black mortality in the summer season (row 9). Klepp includes data for the New England area as well. She estimates that blacks had an excess mortality of over 100 percent above that of whites in Boston during the middle of the eighteenth century. From these data the absence of African slavery as a significant economic institution in the New England colonies of British North America is explicable—in regions with cold climates, cold-weather diseases predominated and Africans died at disproportionate rates relative to Europeans. Warren’s (1997) mortality data for the nineteenth century are for: Providence, Baltimore, and Charleston, and are consistent with Klepp’s conclusions concerning regional differences in the mortality rates of blacks and whites. Warren’s estimates indicate that blacks died at a rate that was 70 percent higher than the white rate in Providence; they died at a 34 percent higher rate than whites in Baltimore; and, despite blacks’ obviously lower income, lower standard of living, and bias against them in the provision of public services in the antebellum South, they died at an (essentially) identical rate as whites in Charleston.

Figure 5.3 illustrates the impact of the Old World migrations on the disease ecology in the colder regions of the New World. Again, starting from the top, Europeans’ demand for agricultural labor in the colonies leads to migrations of Europeans and importations of African slaves. With European settlers come measles, whooping cough, pleurisy/influenza, other lung infections, and in general, cold-weather diseases. With chance and sufficient time, these Old World cold-weather diseases become endemic to the northern regions of colonial British North America. Because they lack both innate and acquired immunities to cold-weather diseases, people of African ancestry contract relatively more of these diseases. Consequently, they are more sickly and die at greater rates than people of northwestern European ancestry. The relatively greater illnesses and mortality among Africans reduce their productivity and the demand for them as agricultural labor. This increases the demand for Europeans as agricultural labor. As a result, there is an increase in the migration of European indentured servants. This feedback restarts the cycle, leading to a decline in the number of Africans and their descendants living in the northern colonies of the British American mainland.

McCusker and Menard (1985, p. 222) provide estimates of the number of African-Americans living in the British mainland colonies that are consistent with our disease story. They report that the percentage of African-Americans in colonial New England stabilized at about 2 percent of the total population (the percentage in 1700 was 1.8, the percentage in 1780 was 2.0); in the middle colonies, it stabilized at about 6 percent (6.8 percent in 1700 and 5.9 percent in 1780). The percentage of African-Americans in the South did not stabilize: in the upper South it grew from 13.1 percent in 1700 to 38.6 percent in 1780 and in the lower South it grew from 17.6 percent to 41.2 percent from 1700 to 1780.

We can now illustrate the economic case for the relative absence of Africans in the northern colonies. Klepp's (1994, p. 477, n. 13) crude death rate data for Boston for 1725 to 1744 of 80 deaths per 1,000 for blacks and 32 deaths per 1,000 for whites, imply a life expectancy of 12 years for blacks and a life expectancy of 29 years for whites when the data are truncated at 100 years. An indenture contract was considerably less than the life expectancy of an indentured European, it usually ranged from four to seven years. Recall from the discussion in chapter 4 that the exact length of an indenture contract was a function of age, gender, location, and skills, where an illiterate 18-year-old male bound for Pennsylvania could expect to serve five years. Because reliable data for indentured servants bound for New England are not available, we chose to utilize data about servants for the northern-most colony for which reliable data exist for the early eighteenth century. We use the price of an illiterate 18-year-old male because that age and skill category would represent the appropriate comparison to the typical imported African slave.

As noted earlier, colonial interest rates are not known with precision, but a good estimate for relatively safe mortgages is 8 percent. Homer and Sylla (1991, pp. 274–75) indicate that colonies passed maximum interest rate laws ranging from 5 percent in Virginia to 8 percent in Massachusetts, noting that “While such legislation does not tell what the prevailing rates were in the colonies, it does indicate the rates that leading citizens considered normal or reasonable. According to Benjamin Franklin, commercial interest rates in Pennsylvania in the latter half of the eighteenth century were between 6 and 10 percent.” We take 8 percent as an appropriate rate, being the midpoint of Franklin's range. Also recall from chapter 4 that we estimated the delivered price of the illiterate 18-year-old male servant to be £12, and here we estimate the purchase price of a slave during the early eighteenth century (1725 to 1744) to be £26. (For the years 1723/27 to 1743/47, the unweighted mean of the slave prices presented in table 4.1, column 2, in chapter 4 is £26.06.) We note that the £26 price used here is well below the slave prices that prevailed during the second half of the eighteenth century. But recall that if we used higher slave prices, that would make the choice of a servant more profitable, other factors the same.

Using these estimates, we can calculate the expected annual returns over costs for European servants and African slaves, given the same productivities, as agricultural labor in the northern colonies. Given the market price of bound servile labor would equal the present value of the expected stream of net revenue produced by the servile labor, we calculate that a servant would have to return only £3.005 per year while a slave would have to yield £3.45 per year (or 15 percent more) for its entire expected life to justify the higher purchase price.15 Recall that people of African ancestry were sicklier in the cold-weather disease environment of the North. These calculations favor purchasing a European servant in the northern colonies; servants did not require as high a yield as African slaves to be profitable. Risk aversion also would favor servants over slaves because the initial investment was less than the price of a slave, and the risk of death was less with a servant. Under these circumstances, we would not expect African slaves to be competitive with European servants in the North, and they were not.

While these estimates do not address the issue of the absolute profitability of indentured servants or slaves, they do suggest that investments in European indentured servants were more profitable than investments in African slaves in the northern colonies. To calculate the absolute profitability of indentured servants and slaves, we must know (1) the acquisition cost of indentured servants and slaves, (2) the rate at which the investment in each is being discounted, (3) the length of time the investment in each lasts, and (4) the net revenues generated by each type of servile labor. All the data are known except for the last. And the argument in this book is that the net revenues generated by African slaves in agriculture differed regionally from those of European indentured servants.

Another indication of the relative profitability of African and European servile labor across regions of the mainland colonies is to estimate the annual imports of each type of labor into each region. The data assembled by Perkins (1980, p. 154) show the regional distribution of slaves and indentured servants in 1774. While the total stock of slaves exceeded the number of indentured servants in the Middle and New England colonies, the annual number of imports of indentured servants substantially exceeded the annual imports of slaves into these colonies.

With a few assumptions we are able to calculate annual imports. We assume that the expected life of a slave was 12 years and the length of an indentured servant contract was 5 years. We also assume that the natural growth rates of the servant and slave stocks were zero. Indentured servants obviously did not reproduce as indentured servants; and an assumption of no reproduction on the part of slaves biases the estimated imports of slaves upward because some did reproduce. In the Middle colonies (New York, New Jersey, Pennsylvania, and Delaware), the stock of slaves was 34,172 and that of indentured servants 21,374. The annual number of imported servants, under our assumptions, would have been 4,275 (1/5 of 21,374) and the number of imported slaves 2,848 (1/12 of 34,172).

Similar calculations for New England (Connecticut, New Hampshire, Massachusetts, and Rhode Island) reveal that annual imports of indentured servants would have been 2,371 and imports of slaves would have been 1,137. The estimates indicate that for each slave imported into New England a little over two indentured servants were imported, and in the Middle colonies the ratio of indentured servant to slave imports was a little over one and a half to one. As one would expect, the calculations suggest completely different outcomes in the southern colonies (Georgia, Maryland, North Carolina, South Carolina, and Virginia). Under the same assumptions, albeit biased upward for slave imports, the data suggest that for each indentured servant imported there were a little over nine slaves imported.

In regions with cold climates, investments in white indentured servants, given the same productivities, were more profitable than were black slaves at the prices that prevailed in the eighteenth century. We expect that the assumption of equal productivities in cold climates to be invalid. Given the disparity in death rates and the nature of many cold-weather diseases to linger (causing high morbidity), it is unlikely that in agriculture Africans were as productive as Europeans in the northern colonies. The disease environment almost precludes that as a possibility.

In a related issue, Hanes (1996) argues that slaves were preferred for employment in personal service (domestic servants) because people used in such service had to be aware of the idiosyncrasies of their masters. Since slaves were permanent (as long as they lived), training costs were reduced by using slaves rather than indentured servants. Personal servants perpetually bound (slaves) would have a greater knowledge of their overlords various tastes and idiosyncrasies; this would have made them more valuable than a constantly changing stock of free labor. As a result, the demand for personal servants may explain the continuing (small) demand for slaves in the North. The slavery that did exist in late-colonial New England and New York was largely confined to domestics with the exception of New York City and parts of Rhode Island. Also there is some literary evidence that African slaves in the North were employed more in personal service than their counterparts in the southern colonies. Nonetheless, our argument addresses the use of labor of different ethnicities (ancestral heritages) in the agricultural sector, not their use as domestic servants.

The British Colonial Mainland: The South

In the tobacco-growing regions of the Chesapeake and into the Carolinas, the environment is more favorable to the transmission of warm-weather diseases. These diseases, their seasonal virulence, and the significant differential impact on Africans and Europeans explain the eventual predominance of Africans as the agricultural labor force in these regions. We concentrate on the Chesapeake region for three reasons. (1) The data and evidence that exist are better for this region than any other region where African servile labor was a major force. (2) The reason for the transition from European to African servile labor for agricultural tasks in this region has been widely studied and has a substantial literature. And (3) in the Chesapeake, there is a very interesting anomaly—the evidence indicates that in the early years in the Chesapeake region mortality rates for blacks were greater than those for whites (Fogel 1989; Klepp 1994). Fogel (1989, p. 125) actually reports a negative natural growth rate for blacks in the early eighteenth century, while whites had a positive natural increase.

The issue then is to explain the existence and evolutionary success of African labor in the period 1680 to 1720 in the Chesapeake region in the face of appreciably higher death rates for blacks. Our explanation involves the seasonality of diseases, the differences between rates of morbidity and mortality, and the spread of hookworm and malaria.

In the agricultural cycle any illness that strikes the labor force during the late summer or early fall harvest season can have a devastating effect on the economic well-being of the agriculturalist. Because the environment favors warm-weather diseases as one travels further south in the Northern Hemisphere, the ethnic differences in seasonal mortality rates would have been as great in the Chesapeake as in Philadelphia. Consequently, part of the conundrum of choosing Africans over Europeans despite the lower life expectancies of Africans can be explained by the seasonality of disease in the South. African-Americans got sick (and died) disproportionately in the winter months—the slack time in agriculture—while European-Americans were brought down in the summer months—at the peak of the agricultural cycle. The economic productivity of Africans relative to Europeans in agricultural pursuits in British North America was greater in the summer and early fall because of the regional predominance of warm-weather diseases.

A number of parasitic diseases afflicted the American South. The climate allowed warm-weather diseases that were endemic to tropical Africa to flourish in the long Chesapeake summers. Some of these diseases preexisted, and others became serious health problems only after Contact with the Old World. Over time some Old World diseases became endemic to the southern disease ecology and made the South disease ridden to people of European ancestry relative to the more northern regions of America. Warm-weather diseases were a major factor in the southern disease environment that affected the economy and differentially affected whites during the slave era. Of these, the primary ones are malaria, yellow fever, dengue fever, hookworm, and schistosomiasis. All of these were (and are) found in tropical West Africa and were widespread and abundant. The climate and sandy soils of the South, as well as the plantation system, were relatively conducive to the maintenance and spread of many infectious parasitic diseases. There were other such diseases that are not known to have differential effects on ethnic groups that also thrived in the South because of its climate and soils; some of these include amoebic dysentery and the helminthes, roundworm, threadworm, whipworm, and tapeworm. The South also had the ubiquitous smallpox.

The best documented and studied of the diseases that had disparate impacts on peoples of African and European ancestry are hookworm, malaria, and yellow fever. Both hookworm and malaria have a low case mortality rate but cause a great deal of sickness. Both are of considerable importance in explaining the ultimate dominance of African slavery in the Chesapeake and, by extension, the entire American South. As for yellow fever, because its vector, the mosquito Aedes aegypti, thrives in centers of very dense human populations compared to the malaria mosquito Anopheles genus, it had little impact on agricultural field labor in the Chesapeake. And, except for the occasional epidemics in the cities that did exist in colonial North America (New York in 1688, Philadelphia and Charleston in 1690, Boston in 1691, and Philadelphia in 1762), yellow fever was not a major factor in the demographic history of the colonial South.

Natives of tropical West Africa tend to have both a genetic resistance and acquired childhood antibodies to the various forms of malaria, while the natives of northwestern Europe lack these bodily defenses. This is not entirely correct because the “ague” or the vivax type of malaria (Plasmodium vivax) was present in seventeenth- and eighteenth-century Europe (Kiple and King 1981, pp. 12–23). This may have allowed some eighteenth-century Europeans to have acquired resistance to strains of Plasmodium vivax, but they did not have an innate (genetic) resistance to malaria. Innate immunities are caused by differential reproductive success; only diseases that selectively kill elements of a population have major effects on the populations’ gene pool. P. vivax is very infrequently fatal, while P. falciparum is the most lethal of the malarial diseases. P. falciparum was never endemic to Britain; consequently, the pressures of evolutionary selection for resistance to malaria did not exist in Britain in the millennia before the modern era. Furthermore, Kiple and King (1981, p. 16) note that any acquired immunities would be limited to a specific strain of a variety of malaria because different varieties and strains, of which there are many, are not immunologically similar.

When Africans were brought to the Chesapeake in substantial numbers in the late seventeenth century, unwittingly malaria of the P. falciparum variety was also imported. An individual may be infected with the malaria plasmodia but, for all practical purposes, be asymptomatic. Yet these infected people carry within themselves fully virulent malaria. (The African forms of malaria are, primarily, Plasmodium falciparum, and secondarily, Plasmodium malariae.) The natives of tropical West Africa are likely (relative to non–West Africans) to have the sickling trait. This is a probabilistic statement. The probability of a West African carrying the sickle-cell trait approaches 40 percent in some populations (Dunn 1993, pp. 855–62; Kiple and King 1981, p. 17). The sickling trait allows the human body to exist and function almost normally while bearing the plasmodia parasites. In tropical West Africa malaria is endemic, people born and raised there have, very likely, been exposed to many strains of P. falciparum by puberty. Some natives of tropical West Africa could have escaped being bitten by malarial- bearing mosquitoes, but the probability would not be high, and their numbers would be statistically and economically insignificant. While the Africans brought to the Chesapeake region could have been inured to the ravages of malaria by the sickling trait, other blood abnormalities, or previous exposures, they still could serve as reservoirs of the disease that allowed the mosquitoes to transmit it to others.16

In the summer, mosquitoes were and are ubiquitous in the Chesapeake region. Mosquitoes transmitted malaria from Africans to Europeans and the Europeans took sick; some died, but the case mortality rate for malaria was low. We do not distinguish between the various types of malaria here because pre–twentieth-century medicine could not accurately distinguish between them. And the available data typically make no mention of the varieties of malaria. While data on morbidity and mortality for malaria in the Chesapeake for the 1680 to 1720 period do not exist, Duffy (1988) has a general discussion of malaria morbidity in the American South. And Curtin (1989, pp. 174–77) reports the causes of morbidity and mortality for British (European) troops in the British West Indies during the nineteenth and early twentieth centuries. Curtin's (1989) data indicate that the case mortality rate for malaria in the Caribbean was approximately 1.6 percent.17 Furthermore, Curtin (1989, p. 133) states that prior to 1840 the case mortality rate for malaria in Madras, India, was 60 percent higher than after 1840. Suppose it also was 60 percent higher in the Western Hemisphere. That implies the pre-1840 case mortality rate for malaria would have been slightly more than 2.5 percent in the Caribbean. (We multiplied 1.58—Curtin’s actual mean percentage of case mortality—by 1.6 to get 2.53 percent.)

The case mortality rate for Europeans in the Caribbean was probably higher than the case mortality rate for Europeans in the Chesapeake because malaria was more prevalent in the Caribbean and it was a year-long disease, not a seasonal disease as it was in the Chesapeake. The number of and different types of plasmodia protozoa in the victim's blood stream determine the case mortality for malaria. Given that malaria was more prevalent and year round, a victim in the Caribbean was more likely to be bitten by greater numbers of infectious mosquitoes carrying a greater variety of plasmodia types than a victim in the Chesapeake.

Although 2.5 percent is likely to be an upper bound estimate of the malaria case mortality rate in the Chesapeake for the 1680 to 1720 period, how probable is it? Suppose the morbidity rate for malaria in the Chesapeake was 200 cases per thousand. (This also is likely to be an upper bound estimate; Curtin's 1989, pp. 174–80, morbidity rate for European troops in the Caribbean ranges from over 300 cases per thousand to 90 cases per thousand.) The arithmetic implies that malaria could have contributed, at most, an additional five deaths per thousand (2.5 percent of 200) to the European and European-American death rate. Walsh and Menard (1974, p. 224) estimate that, for the late seventeenth century, males at age 20 in New England had a life expectancy of between 10 and 26 more years than males in Maryland (the Chesapeake). An additional five more deaths per thousand does not explain this gap. Accordingly, the evidence from the death rate data that do exist makes estimates of a case mortality rate of 2.5 percent and a morbidity rate of 200 per thousand for malaria in the Chesapeake more than plausible. (Rutman and Rutman 1976 have a similar accounting in their study.) The estimates are more than plausible because, given what we know about the disease ecologies of the two regions (New England and the Chesapeake) and how they changed, it would be implausible if malaria alone accounted for all or most of the gap. The estimates suggest that malaria could have been doing a lot of damage to the productivity of whites in the Chesapeake (high morbidity) but still not make a dent in the difference in crude death rates across regions (low case mortality). Consequently, to quantitative economic historians, who often insist, “If it can't be counted, it didn't exist,” the impact of malaria might be invisible if mortality rates are the only data included in the metric.

The history of hookworm in America is more obscure than that of malaria. Hookworm was not recognized as a major problem in the United States until 1902 when Charles Wardell Stiles conducted microscopic studies on hookworm among the southern population and found it to be endemic in the South. Hookworm is an infestation of intestinal nematodes that is a seasonal, warm-weather disease that parasitize humans. The type of hookworm infecting the American South is infrequently fatal, but it can be debilitating. How debilitating hookworm is depends on a variety of factors: general health, nutrition, physical activity, and parasite load. The same parasite load in a poorly nourished population that engages in heavy physical exertions, compared to a well nourished one, can have severe effects; on the poorly nourished population, these effects include chronic sluggishness, weakness, exhaustion, and, in extreme cases, even cerebral hemorrhage.

Hookworm disease has ethnically (racially) disparate effects: natives of and descendants of people from tropical West Africa resist it better and tolerate a given parasite load better than do natives of and descendants of people from northwestern Europe. The exact scientific reason that people of tropical West African ancestry have greater resistance to hookworm is unclear. That they do was recognized almost as early as the disease (Stiles 1909). Indeed, in one study of descendants of people from tropical West Africa residing in the Caribbean, it was found that people infected with hookworm had higher productivity than the non-infected (Weisbrod, Andreano, Baldwin, et al. 1973, p. 76).

The impact of hookworm on people of northwestern European ancestry is different. Brinkley’s (1994, pp. 113–14) findings suggest that in areas in the American South with ideal conditions for the propagation of the hookworm nematode, the percentage of whites in the area had a significant and negative impact on per capita agricultural income in 1910. His findings also indicate that locations that likely had heavy hookworm infestation in the post–Civil War South—an area in which people of African ancestry were systematically discriminated against—were places where “A county composed of 100 percent blacks would have a higher income of between $10.50 and $23.80 than a county of 100 percent whites, ceteris paribus” (Brinkley 1994, p. 114). Brinkley (1994, p. 84) also cites estimates by medical doctors of hookworm's affect on the reduction in early twentieth-century agriculture productivity that range between 20 and 70 percent with 40 percent being the most common estimate (see also Brinkley 1995, 1997). The exact reduction in the productivity of agricultural workers, however, is not the central issue here, what we do know is that hookworm reduced productivity differentially: people of European ancestry suffered more severely than people of African ancestry. There is every reason to believe that this ethnic disparity also existed in the centuries that preceded the twentieth century.18

We can summarize our explanation for the transition to African agricultural field workers in the Chesapeake with the help of figure 5.4, which is a combination of figures 5.2 and 5.3. We begin with the importation of African slaves into the Chesapeake in the seventeenth century (the top box in figure 5.4). Africans carry hookworm, malaria, and other tropical pathogens, within themselves. After the number of Africans increase beyond a threshold level, tropical diseases become endemic. Mosquitoes then transmit the malaria to the Europeans, and the hookworm nematode finds hosts among the Europeans as do other tropical pathogens. The Europeans are relatively sickly and die (not very probable), or are sickly (highly probable) during warm weather and the harvest seasons when tropical pathogens flourish. The relatively greater illness among Europeans would decrease the supply of those willing to voluntarily indenture themselves as agricultural field hands in the Chesapeake region (the bottom right-hand side box in figure 5.4). The illness among Europeans likewise decreases their agricultural productivity, which reduces the demand for Europeans as agricultural field hands. This increases the demand for Africans, leading to increased importation of African slaves, which restarts the cycle and leads to a relative decline in the number of Europeans and their descendants living in the southern colonies of the British American mainland.

In the British North American colonies of South Carolina and Georgia, hookworm and malaria were similarly associated with the introduction of African slavery. In South Carolina, Childs (1940, p. 263) argues that with the regular importation of African slaves, malaria had become endemic by the late 1690s; Wood (1975, pp. 62–69) notes that high mortality rates among whites discouraged their immigration to South Carolina at the end of the seventeenth century. Georgia, initially established as a free colony (where slavery was prohibited), did not become malarious until after its charter was altered in 1752 and slavery was established. Cates (1980, p. 153) writes: “Rapid expansion of slavery in the 1750s and 1760s [in Georgia] introduced large numbers of malaria infected Negroes to the colony. Thus, the very changes that most colonists felt were necessary to a prosperous and vital society produced an environment favorable to the growth and dispersal of parasites that would prey on it.” Hookworm and malaria transformed the American South; once introduced, they became endemic. Given the knowledge of the time, these changes were irreversible.

Concluding Thoughts on the Initial Location of Africans and Europeans

History is written from an anthropocentric viewpoint. As humans, this should not cause us much concern except when nonhuman organisms materially impinge on the course of human events. The discovery of the New World and the interchanges of biological organisms (human and nonhuman) was one of those events. Prior to Contact, the New World was relatively disease free. Immediately after Contact, the disease environments of its different climatic and geographic regions began the process of transformation into mirror images of their Old World counterparts. The African slave trade opened the New World to Africans and the diseases of tropical West Africa. This affected the patterns of settlement in the New World.

The value of servile labor depended on the discounted stream of income produced. In the Caribbean, the stream of earnings produced by an African was, on average, at least twice as long as that of a European because of the African's greater life expectancy. The European's income stream in turn was markedly longer than that of a Native American. The African also was likely to be healthier. Unless native or European servile labor was substantially cheaper, or their productivity in the same occupation was substantially greater than their African counterparts, neither would survive economically in competition with Africans in the New World tropics.

The biological susceptibility of northwestern Europeans to the disease environment that eventually prevailed in the New World tropics substantially reduced their life expectancy and capability for work. Given the alternative of enslaved Africans, European labor, servile or free, could not be supplied at a price low enough to be a viable alternative. The native populations fared even worse. The reason for these ethnic differences is that the disease environment that eventually prevailed in the tropics more closely resembled that of tropical West Africa rather than that of northwestern Europe or the pre–Contact New World. For millennia generations of West Africans had been exposed to virtually all tropical diseases. West African peoples had immune systems that had been selectively culled by the environment for the greatest resistance to tropical pathogens. The surviving and reproducing Africans of tropical West Africa produced a population that was, relative to Europeans, much more resistant to the onslaught of tropical diseases.

On the mainland of the North American continent, the environment became less favorable for tropical pathogens as one moved north. Cold-weather diseases were sufficient to make the northern colonies relatively unattractive for African slavery. In the Chesapeake, a combination of warm-weather diseases and agricultural seasonality were sufficient to make African slavery thrive. Further south, the introduction of African slaves changed the disease environment so that African slavery was the most economic form of labor given the institutions that existed.

How well does this disease story fit with the existing historical chronology? Economic historians have frequently written about the transition from European indentured servants to African slaves in British North America. Their explanations are driven by relative price differences: In the late seventeenth and early eighteenth centuries, the price of indentured servants rose relative to the price of slaves. Our explanation is perfectly consistent with this: a relative price shock induces the large-scale importation of African slaves. The timing of the price shock coincides with and is at least partially explained by the ending of the Royal African Company's monopoly of the English slave trade and a series of European wars that disrupted the markets for both African slaves and European indentured servants. The wars made African slaves cheaper to the English by disrupting the French and Spanish slave trades, and simultaneously raised the price of unskilled Europeans to the New World settlers because the wars increased the demand for soldiers, and thus raised the price for young, unskilled males—the type of person most likely to become an indentured servant (Grubb and Stitt 1994).

But we tell the remainder of the story, and resolve the dilemma of why Chesapeake plantation owners did not switch back to indentured servants when the price of African slaves rose relative to Europeans later in the eighteenth century, and why slavery was not viable in the northern regions. Once tropical diseases became endemic to the warm-weather areas of the New World, the economic calculus changed: Europeans were no longer good substitutes for Africans in agricultural field labor. Repeating a point previously made, these effects were irreversible and thus path dependent given scientific knowledge at the time.

Past and present historians have noted the “unsuitability” of European labor in the tropics and subtropics of the New World. In contrast, economic historians have made the (often implicit) simplifying assumption that Africans and Europeans were similarly productive and thus substitutable. The assumption of (approximately) homogeneous labor led economic historians to concentrate on changes in the supply of indentured servants and slaves in explaining the switch to African slaves in the New World. What we do is call attention to the differential reaction of peoples of different ancestral origins to the Old World pathogens introduced into the New World that led to changes in the demand for servants and slaves. A somewhat paradoxical result is that African slavery was a curse to not only the enslaved African, but also to the Europeans (and their descendants) in the New World where the environment was conducive for the growth and transmission of warm-weather diseases. Only after a significant African slave trade was established did the British American New World disease environment begin to resemble that of tropical West Africa. The changed environment made the American South a pest house, and the New World’s tropics a graveyard for Europeans.