Early Achievement of Modern Growth:
Height and Health of Free Black Children in Antebellum Virginia

Howard Bodenhorn
Department of Economics
Lafayette College
Easton PA 18042-1776
Telephone: 610-250-5306 Fax: 610-250-8961

Although efforts to link trends in human stature to trends in economic well-being remain contentious, the link between height and economic inequality is well established. Human biologists, in fact, suggest that cross-sections are the real strength of anthropometric data because it readily exposes class differences in health and nutrition. Using information on modern heights, Steckel (1983) found that relatively small increases in income inequality (measured by the Gini coefficient) could offset the stature-enhancing effects of large increases in per capita income. Steckel's result, combined with declines in mean heights, led Fogel (1986, p. 498) to speculate that income inequality rose markedly in late-antebellum America. The American poor were growing relatively poorer.

To students of the African-American experience this seems both plausible and probable. The poorest of the poor were free blacks living in the shadow of slavery and their condition deteriorated in the antebellum era. The historiography of blacks is one of people living in wretched poverty. Their economic and civil liberties were regularly trod upon and they were left with few opportunities for advancement, scarcely enough to eke out a bare subsistence. On the eve of the Civil War, their status had grown so piteous that Berlin (1974) labeled them "slaves without masters."

Anthropometric evidence paints a different picture. Previous studies by Komlos (1992) and Bodenhorn (1998) show that black adults grew nearly as tall as whites and modestly taller than slaves. But the complexities of growth do not imply that similar adult stature translates to similar disease experiences, similar diets or similar growth paths during childhood. We now know that slave children followed a growth path unlike any other known population. Virginia's free black children, on the other hand, followed a distinctly modern growth path, one resembling that of late 20th century African-Americans. Indeed, there was something peculiar about the "peculiar institution." Moreover, Virginia's free blacks slaves without masters. They shared few of the slaves' disadvantages.

The Virginia Registrations and Free Black Heights
In the years following the American Revolution, southern whites grew increasingly concerned with the growing number of free blacks. The Virginia legislature responded with a number of measures including a law requiring them to register with the county clerk. These clerks recorded detailed information about registrants, including their name, age, gender, height, complexion, unique scars or other physical characteristics and whether they were free-born or manumitted. Virginia law presumed that black skin and slavery were concurrent, so it was the black's responsibility to prove his or her freedom when asked, and a copy of a registration provided that evidence. Accuracy and precision were vital which makes these documents a particularly valuable historical resource.

Most blacks registered in their teens and early twenties, but some parents registered their minor children. Information on free-born black children was gathered from the registers of 25 rural counties that include several counties located in the Atlantic coastal plain, the Piedmont, and the Appalachians. The registers yielded information on 2312 free-born black males and 2130 free-born black females ages 1 to 22.

Because older children and young adults are over-represented, raw averages were smoothed using Preece and Baines's (1978) Model I. The estimated height-by-age profiles are reported in Table 1 with sample sizes and centiles of modern height. The first of several notable features of these growth profiles is that black males attained a terminal adult stature (67.5 inches) about one-half inch less than contemporary white males, and about one-half inch greater than male slaves. This immediately suggests that the nutritional status of black males fell somewhere between the white and slave standards. Black women attained about the same stature (62.5 inches) as slave women, but so little is known about the white female heights that conjectures on relative nutritional status would be speculative and premature.

A second notable feature of the black growth profile is that it differed markedly from that reported for slaves. Steckel (1992a, p. 498) noted that the "height of a typical slave child would trigger alarm in a modern pediatrician's office" because, until age 10, they fell below the first centile of modern height. The slaves' subsequent growth, however, was remarkable. Pathologically short slave children experienced a supra-normal late adolescent growth spurt that allowed slave males to ultimately attain the 16th centile of modern stature. Slave women attained the 20th centile. Virginia's black children followed a different growth path. Unlike slaves they entered a period of supra-normal childhood catch-up growth and attained the 10th centile of modern stature at age 10 and nearly the 20th centile at 21 years. Black girls followed a similar growth path. Black youth seemingly faced more hospitable diet and disease environments than that into which slaves were born.

Table 2 highlights the third notable feature of black childhood growth, namely that it followed a path resembling that of modern black children. Between ages 4 to 10 antebellum black males progressed from the 2d to the 10th centile; they then declined to the 2d centile at age 16; they then grew rapidly to attain the 19th centile at maturity. Twentieth-century black males follow a similar pattern. From age 4 to 9 modern black children progress from the 10th (note that modern blacks, too, are relatively short) to the 35th centile; they then decline to age 13 after which they grow rapidly and ultimately attain the 42d centile. Antebellum black males followed a typically modern growth path, albeit with lower peaks and troughs that occurred about one to two years later than modern African-Americans. Antebellum black girls followed a similarly modern growth path. Another notable feature is that antebellum blacks followed growth paths unlike most others of African descent. Table 2 reports centiles of modern stature for several black populations in developing countries and the results show that blacks in less developed countries tend to be relatively tall children but relatively short adults. They often decline from the 40th centile of American stature at ages 4 or 6 to the 10th or 20th centile in their late teens to early twenties.

Taken together, these features suggest that antebellum black children prospered when compared to slaves and peoples of African descent in developing countries. While they almost certainly consumed a less nutritious diet and faced a more virulent disease environment than modern African-Americans, antebellum blacks followed a growth path more reminiscent of first(tm)world twentieth century norms than those of modern developing countries. They grew as tall as modern Somalians and Cubans. They were markedly taller than Ethiopians and Arubans. Virginia's black population does not appear quite so piteous after all.

Comparative Income, Mortality and Morbidity in Antebellum America
Many historians have accepted the premise that southern blacks were a dispirited underclass, portending, perhaps, the South between Reconstruction and Civil Rights. O'Brien (1996a, 1996b), for example, calculated black per capita income only about 53 percent that of whites in 1860. Blacks undoubtedly earned less per capita than whites, but it seems that occupational crowding rather than unequal pay for equal work was the cause. Discrimination manifested itself in low(tm)status occupations for blacks. But while there were few black professional, managers or merchants, black artisans and laborers readily found employment and drew wages comparable to whites.

Although the mapping from income to height is not well specified, the function appears to be nonlinear and, perhaps, sharply discontinuous. Small percentage increases in income generally translate into large changes in height at very low incomes, small effects at modest incomes, and no effect at high incomes. Without knowing anything of the growth profile of antebellum whites, the modernity of the growth path followed by Virginia's free blacks and the distinctly un-modern path followed by slaves suggests that blacks (53 percent of white income) had crossed some critical income threshold that slaves (35 percent) had not (O'Brien 1996a, 1996b).

Steckel (1987, 1992a) focused on two explanations of short slave children: (1) high work intensity among pregnant and lactating slave women, which produced intra-uterine undernutrition and shortened periods of breast-feeding; and (2) poor diets in childhood that resulted in protein(tm)energy malnutrition. Poor nutrition from early fetal development through late childhood produced stunted slave children, a condition not reversed until children were simultaneously put to work and provided with a relative surfeit of food. Coelho and McGuire (1997) provided an alternative explanation for high infant mortality, low birth weights, stunted children and the late adolescent growth spurt. Slave children, unlike rural white and black children, grew up in close quarters with a few to a few dozen coevals. Crowded quarters increased the frequency and, possibly, the intensity of infection. The late-adolescent growth spurt was, then, not the result of improved diet as slave children were handed their first hoe and sent to the field. Rather, it resulted from being transported from the penumbra of diseases encircling the slave nursery.

Like the anthropometric evidence, age-specific mortality statistics reported in Panel A of Table 3 suggest that blacks experienced a childhood more like the one lived by whites than that of slaves. Infant mortality rates among male slaves (159 per 1,000) were nearly twice that of whites (94) and blacks (89). Female slave infant mortality, too, was about twice that of whites and blacks. Mortality rates across all groups declined sharply after age 1, however, and gradually converged, though slave mortality rates slightly exceeded free rates. The census may have under(tm)counted mortality, a possibility that raises two issues: (1) the extent of the undercounting, and (2) whether undercounting, if there was any, differed across races. Mortality estimates from other places and times suggest that undercounting at early ages was not widespread. Slave infant mortality, for instance, is comparable to that of late-nineteenth century Europe and several modern African and Asian countries. Infant mortality rates for the free population are similar to late-nineteenth century Scandinavian countries. Using plantation records, however, Steckel found slave infant and child mortality rates well in excess of those reported above.

But three features differentiated Virginia's slaves from Steckel's. First, Steckel drew his sample from large plantations that had higher mortality rates than small farms. Second, Steckel's sample was drawn predominantly from rice, cotton and sugar plantations where pregnant and lactating slaves worked harder and longer than women on Virginia's tobacco farms. And the harder the mother's work, the lower the neonatal survival rate. Third, Steckel drew his sample from coastal South Carolina and Louisiana, regions with more virulent disease environments than Virginia's Piedmont. For present purposes, then, it is reasonable to believe that the census provides a reasonably accurate comparative portrayal of infant and child mortality.

Panel B provides disease-specific mortality rates. These statistics bring into sharp relief the sources of excess slave mortality. Kiple and Kiple (1977) argued that high slave mortality rates resulted from evolutionary maladaptation among African-Americans. Blacks had evolved to survive in an African climate, on an African diet, facing African diseases. When they were transported to North America, however, they faced a North American climate, ate a North American diet, faced North American diseases and fell victim, in disproportionate numbers, to Darwin's process. Statistics from antebellum Virginia fail to support the Kiples's hypothesis. The likelihood that an individual would contract and die from any number of diseases was significantly higher for slaves. Neonatal tetanus -- usually a result of improper treatment of the umbilicus -- took a huge toll in the slave quarters. Acute respiratory diseases (notably, pneumonia, whooping cough and scarlet fever) were more likely to kill in the slave quarter than in free households. Other diseases, like acute diarrheal diseases, were equal opportunity killers, taking huge tolls on infants and children of both races, free and slave. The principal black-white difference was mortality attributed to intestinal worms. Coelho and McGuire (1997) suggested that helminth infections, like hookworm, were in fact more common in the slave quarters than in the white household, but hookworm would not have been easily recognized in antebellum Virginia. It seems more likely that helminth-related deaths were the result of roundworm, flatworm or tapeworm infections, diagnoses of which do not require microscopic evaluation. Statistics from the Southeast suggest that worms may have been a bane for all blacks, not just slaves. White infants, on the other hand, were more likely than blacks to succumb to malarial fevers. This may be a case where African heritage provided a clear evolutionary advantage, but whether malaria resistance among blacks is inherited or acquired remains unresolved among malariologists.

Although mortality data provides a picture of how people died, it does not necessarily tell us much about the diseases that survivors battled. We can, however, draw on other bits of evidence. One is contained in the 1860 mortality census itself. Census marshals were asked to provide a brief description of the most common diseases (not just those culminating in death) in their county. Consistent with the mortality data, typhoid and pneumonia were the most commonly identified diseases, but scarlet fever, diphtheria, malarial fevers and diarrhea were nearly as frequent. Although typhoid's highly infectious nature made its appearance a matter of intense concern, it was recognized that it need not be deadly and when it was, it was usually as a result of complicating factors. The census marshal for Patrick County, Virginia explained that typhoid was "the most prevalent disease in this county but my opinion is that it invariably yields to proper treatments unless the constitution has been broken down by some imprudence or some other disease before the attack of fever."

What, then, were these complicating infections? Historical information on case-fatality rates and disease prevalence in developing countries with large populations of African heritage provide some clues. Acute diarrheal diseases, for instance, are endemic but rarely kill. Case-fatality rates in Latin America and Sub-Saharan Africa are about 0.1 to 0.2 percent. A typical child experiences 3 to 7 episodes per year with each acute attack lasting approximately one week (Murray and Lopez, 1996). Given the sanitary and hygienic (non)practices of early Americans, similar rates and durations seem likely. Acute and repeated diarrheal infections (which encompasses a wide variety of enteric diseases), especially in combination with dietary restrictions, are widely recognized as a cause of childhood stunting. Recovery or catch-up growth can be complete and rapid if infections are few and widely spaced, but this seems unlikely in the slave quarters where repeated infection was a fact of life. Recurrent acute respiratory diseases (like pneumonia, whooping cough and scarlet fever), too, were a fact of life and each has deleterious effects on both physical and cognitive development. A severe, medically untreated case of whooping cough, for example, results in stunted and cognitively impaired children. One of the many underlying causes of short slave children and (relatively) tall free black children was the differential incidence of respiratory diseases between the two groups. Unlike slave children living in close quarters, black children, like white children, lived and grew in households that little resembled the slave quarters. And, unlike slave children, free black children experienced a pattern of growth resembling that experienced by modern African-Americans.

Just as there are nonlinearities in the mapping of income to height, there seem to be sharply non-linear and discontinuous effects of disease on height and growth. A particular disease or complex of diseases was more likely to generate stunted children when those children lived on the margin of malnutrition. It is in this, perhaps, that we will find a resolution to the diet-or-disease debate. Slave children lived in a unique disease environment. What made it unique, however, was not the type of disease but the greater incidence of disease. Coelho and McGuire (1997) focus on hookworm and malaria, diseases probably transported with slaves to the New World. But the real bane of slave children was the so-called diarrhea-pneumonia complex that the World Health Organization still battles throughout the world. Non-existent hygiene, poor sanitation, sub-standard housing and crowding in the slave quarters promoted the transmission of respiratory and enteric diseases. During an illness, even a mild one, growth slows or, sometimes, altogether ceases. But if nutrition is adequate and the number of pathogens small, this slow-down in growth is followed by a brief period of supra-normal catch-up growth which restores the child to his or her original growth path. But chronic undernutrition and continual reinfection short-circuit the growth recovery response. The result, of course, was severely stunted slave children whose period of remarkable recovery was delayed until they were transported to the fields, where they escaped the cycle of chronic reinfection and undernutrition.

Antebellum Virginia's free black infants, children and youth did not face the same dietary and disease environment as slave children. Eveleth and Tanner (1990, p. 198), among others, contend that children from lower socioeconomic groups are shorter than their better-off coevals. "In consequence the height of children in the various divisions of society -- as well as their tempo of growth -- may be used as a measure of the health and general well-being of the group in question...." Both terminal heights and the pattern of growth of free black children fail to support the premise that they were forced by racism and discriminatory laws into an underclass culture of poverty. There can be no doubt that they were discriminated against and faced various disadvantages, but those disadvantages were not so great that their children were hungry and diseased, at least no more so than most whites. It may be time, then, for a reinterpretation of the free black experience in antebellum America.

Please note that the following tables in the original summary contained formatting that cannot be reproduced accurately in this web version. Hard copies are available from The Cliometric Society on request (csociety@eh.net) or from the author.

Table 1
Estimated Heights of Virginia's Free Blacks Compared to Modern Standards

		           Males              		           Females                   
		Estimated	Sample		Modern		Estimated	Sample		Modern
Age		Height		Size		Centile		Height		Size		Centile
2		31.4		 11		 0.6		30.5		 13		 0.2
3		34.8		 17		 1.1		34.4		 10		 0.8
4		38.0		 23		 1.9		37.9		 19		 2.7
5		40.8		 14		 2.9		40.9		 18		 5.8
6		43.4		 21		 4.6		43.6		 17		10.4
7		45.8		 16		 6.6		46.0		 25		13.4
8		48.0		 23		 8.5		48.1		 16		13.6
9		50.0		 26		 9.5		50.1		 31		12.7
10		52.0		 25		10.2		52.2		 23		10.9
11		53.8		 26		 8.9		54.3		 28		 7.8
12		55.7		 32		 7.8		56.6		 51		 6.1
13		57.6		 36		 5.9		58.9		 56		 8.4
14		59.6		 42		 4.3		60.7		 74		14.9
15		61.6		 48		 2.6		61.8		 88		22.4
16		63.4		 61		 1.9		62.3		149		24.5
17		64.9		 85		 3.4		62.5		155		22.1
18		66.0		118		 7.8		62.6		286		21.2
19		66.7		129		12.7
20		67.1		202		16.4
21		67.4		881		19.2
22		67.5		471
Notes: Heights estimated using procedure outlined in Preece and Baines (1978).  Centiles of modern height calculated from Steckel (1996).
Sources: Virginia register sample.

Table 2
Centiles of Modern Stature Attained by Various People of African Heritage
Table 2 cannot be accurately reproduced and is not available on this web site.
Notes: Underlying heights smoothed using Model I of Preece and Baines (1978). Centiles calculated from Steckel (1996).
Sources: Slaves from Steckel (1987); Modern U.S. blacks and Jamaicans from Eveleth and Tanner (1990); black Cubans from Laska-Mierzejewska (1970); Arubans from van Wering (1981); Somalians from Gallo and Mestriner (1980); Ethiopians from Dellaportas (1969).

Table 3
Child Mortality Rates by Age and Cause
Table 3 cannot be accurately reproduced.

Table 3
Child Mortality Rates by Age and Cause
Panel A: Age-Specific Child Mortality Rates (per 1,000 live births)
                     Whites           Blacks	 Slaves
               Male female     male female  male female
< 1 year        94     71      89    75        159    137

1 - 4 years    21     17      20    13        29     26

5 - 9 years     5      5         6     6           7       9

10 - 14 years  2      3         4     3          5       6
Panel B: Disease-Specific Mortality by Race and Age 
(per 100,000 population)
Age         Whites     Blacks  Slaves

Acute Diarrheal Diseases
< 1 yr      986.2    742.1    827.3
1 - 4 yrs 267.2    233.9   187.5
5 - 14 yrs   26.3       26.6     48.8

Acute Respiratory Infections
< 1 yr        1,520.4     1,793.4    2,412.2
1 - 4 yrs      803.2        643.3    891.5
5 - 14 yrs    191.7        173.0    248.9

Gastro-intestinal Complaints
 < 1 yr        135.9      123.7    252.6
1 - 4 yrs      56.0       0.0      126.4
5 - 14 yrs    16.1        6.7      15.6

Malarial Fevers
 < 1 yr        503.5     309.2   322.2
1 - 4 yrs    104.0     146.2   93.7
 5 - 14 yrs    20.1      39.9   20.6

Intestinal Worms
    < 1 yr         17.3         61.8    69.7
    1 - 4 yrs        25.6            146.2    98.1
    5 - 14 yrs      3.9         13.3    15.6

    < 1 yr        624.3        618.4    1,611.1
    1 - 4 yrs        99.2            87.7        87.2
    5 - 14 yrs     10.2         33.3    30.6
Source: 1860 Mortality census manuscripts for Virginia, North Carolina, Tennessee and Kentucky