Wilbur Zelinsky
“The Population Geography of the Free Negro
in Ante-Bellum America

Population Studies, Vol. 3, No. 4 (Mar., 1950), pp. 386-401

In the years before the Civil War the free Negro formed a numerically minor but socially significant segment of the American population. Although the anomalous position of this group, who dwelt in a limbo between slavery and citizenship, was accompanied by demographic and distributional traits that were quite unlike those of other groups, there has not yet been any general treatment of these topics in the small literature on the free Negro.' My attention was directed toward this investigation by the interesting differences in distribution that have existed between slave and free coloured populations in Latin America 2. In Cuba and Puerto Rico these discrepancies are well documented by census materials even though they are in many respects still unexplained, and they doubtlessly occurred elsewhere in Latin America, and particularly among the large Negro populations of Brazil, where data are lacking or deficient. Although this study is intended as a contribution to the completer understanding of American historical geography, it is hoped that it may also be of value to the scholar who will ultimately undertake a definitive monograph on the free Negroes of the United States. The final and perhaps unattainable aim of this line of research is to discover whatever principles may have underlain the spatial differentiation of the two Negro communities-slave and free-throughout the Americas with the hope of adding to the basic store of population theory.


Before the group was submerged in the mass of emancipated slaves at the conclusion of the Civil War, the number of free Negroes was enlarged steadily from several sources. By far the greatest number were manumitted slaves or their descendants. Manumission, or some related form of legal declaration of freedom, was conferred on favoured individuals for meritorious services or because of sentimental or moral reasons with some frequency all through the history of American slavery, but the persons thus liberated were by no means a representative sample of the Negro population. Slaves who were permitted to hire themselves out as labourers were sometimes able to save sufficient money to purchase their freedom from their masters. A significant number gained freedom by escaping their owners and isolating themselves in remote localities in the South or by fleeing to free soil in the North or in Canada. Quite unimportant, except in a few of the major ports, were those free Negroes who were immigrants from foreign lands. The free Negro population of any given area in the United States would be affected by the more or less legal migration of individuals who had been liberated by legitimate means; and, finally, the annexation of such territories as Louisiana, Florida and Texas added to the size of the group within the borders of the United States. The sum of individuals thus derived would also, of course, experience a natural increase with the offspring not only of members of the free Negro community but also from unions between free Negroes and non-Negroes. Only slightly counterbalancing this continuous increase were three processes whereby individuals were removed from the free Negro population. Re-enslavement was a threat that faced many who remained in the slave states; there was a small trickle of emigrants to Africa and other foreign parts; and very light mulattoes were, in some instances, able to 'pass' into the white population and lose their negroid status. The social, legal and economic status of the free Negro has been treated exhaustively in essays concerned with particular states and cities. It need only be said here that these former slaves and their progeny suffered many disabilities and as a group had not greatly advanced above the slave population. Nevertheless, their very existence, combined with their occasional ability to make themselves felt politically and economically, did make them the object of unceasing interest, suspicion and often hostility in a land where Negroes were generally considered a race of slaves to be insulated from all aspirations to freedom.

Figure 1

Population characteristics

In their demographic characteristics the free Negroes differed greatly from the slave and particularly from the white population of the United States. First of all, they were physically distinct as a group from all other elements in the American population. It is practically a truism that there was considerably more admixture of white blood in the free coloured than in the slave population. The 1860 census states that 36 22 % of the former but only 10.41 % of the latter were classified as mulatto. Although this fact is irrefutable evidence of a manumission differential favouring the offspring of white-Negro alliances, because of a feeling of moral responsibility, it must be remembered that mulattoes enjoyed easier access to other avenues of freedom that had no sexual connotations. A relatively great amount of Indian blood in the free Negro population may also be taken for granted 3.

The free Negro was more frequently a city dweller than either the slave or the white American. Using the somewhat defective data on residence offered by the census of 1860, we find that 2-79 % of the total urban population were free Negroes even though they formed but 1.55 % of the aggregate population. To put it another way, there were 79 % more free Negroes in the cities of America in 1860 than would be expected if the group were uniformly distributed. This is hardly surprising in view of the fact that economic opportunities were much superior for landless and ownerless Negroes in cities than they were in the countrvside. The severe social isolation that must have been encountered in rural areas could have been a potent factor pushing individuals and families toward relative social normality in sizeable urban communities of free Negroes. There is also the possibility that urban slaves enjoyed a better opportunity for manumission or self-purchase. Only in Delaware and New Jersey, states with large cohorts of free coloured inhabitants, did the rural individual surpass the city dweller in importance in 1860. This would indicate that only there and in Maryland, where parity was nearly achieved, was there a real free Negro yeomanry with strong roots in the countryside. The rural element is moderately well represented in New York, southern New England, Virginia and North Carolina, which with the aforementioned states comprise nearly all of the great population hearth of the free Negro. It is in the Deep South and most of the Middle West that the rural individual was relatively rare; it can be suggested that urbanism was inversely proportional to population size; the sparser his representation in a region the more likely was the free Negro to be found in a city.

Figure 2

The sex ratio of the free Negro population was abnormal to a marked degree. The figure of 92-2 for 1860 contrasts sharply with 100.6 for the slave population or 105.2 for the white population. Inasmuch as no reference to vital statistics could explain the wide gap between free Negro and slave sex ratios, and since there existed an excess of adult Negroes, especially young females 20 years and older and the aged of both sexes, we are required to conclude that female slaves not only enjoyed the high rate of manumission accorded ancient retainers but that young women were freed as a token of favour. This distinctly female cast of the free Negro population was in all probability one of the by-products of widespread illicit relations between masters and slaves.

As has been implied already, the age composition of the free Negro community was distinctly abnormal. Children were under-represented, while the mature and especially the aged were relatively quite numerous. Birth statistics for this early period are too crude to permit any definite statement as to the fertility of the group, but one may hazard the guess that the strong emphasis on urban living must have had as an inevitable consequence a lowered birth-rate. In large part, however, the skewness of the age-composition curves may have resulted from the predominantly adult ages of those slaves who were being freed or who escaped servitude.

Population history and distribution

In colonial times the free Negro population was, as is indicated in a quite imperfect record, small and in many places altogether insignificant. Certainly, prior to 1790, the data are too fragmentary and unreliable to allow cartographic synthesis. Beginning with the first census, the time series of maps (Figs. 1-4) takes us from 1790 to 1860, the period of the most rapid growth and greatest demographic significance for the free Negro. Although the census data have been plotted without any attempt at adjustment, it must be remembered that they can be accepted only with serious reservations. The free Negro was undoubtedly considerably underenumerated in both North and South, and this was particularly true for fugitives and for those living in remote localities. Table i summarizes the progress of population growth in the years with which we are concerned.

The picture presented for 1790 is relatively simple as regards both absolute and relative importance. Four principal nodes of concentration existed: southern New England, with a surprisingly dense cluster in Rhode Island; New York City and the Hudson Valley; south-eastern Pennsylvania, Delaware and eastern Maryland; and south-east Virginia and north-east North Carolina. Free Negroes were scattered throughout the Piedmont region, but on the frontier, except in the Great Valley of Tennessee, they were rare. New York was the leading urban centre, followed by Philadelphia, while all other cities were relatively unimportant. There are grounds for supposing little long-distance migration at this time inasmuch as slavery was practised almost universally in the United States. The New England population and others north of the Mason-Dixon Line may be assumed to have originated from local slaves; Rhode Island and New York State were the principal northern centres of slavery during this period.

Figure 3

During the next 20 years the free Negro population virtually doubled in size, and its distribution took on most of the characteristics that were to distinguish it until the Civil War. The old population centres had added considerably to their numbers, with the whole Chesapeake Bay area and the Eastern Shore Peninsula especially well supplied with free Negroes. The urban centres had waxed mightily, and Washington and Baltimore, which was surprisingly inconsequential in 1790, were coming up rapidly as competitors with New York and Philadelphia. A definite concentration of population appears in south-western Pennsylvania and had begun to develop in south-western Ohio; the Blue Grass Region and the Nashville Basin emerge quite clearly; the free Negro had reached areas near the frontier. In the south the Piedmont habitat was favoured and the coastal plain shunned south of Pamlico Sound except for Charleston and Savannah. The Louisiana Purchase had added two highly interesting contingents of free Negroes to the United States: those in the French settlements of Missouri and the large New Orleans population, together with a rather small rural group in the deltaic hinterland.

After 1810 the outlines of the pattern discerned in that year were filled in more clearly. The major part of the population was still clustered on or near the coast between Boston and north-eastern North Carolina in the map for 1830, but this single grand agglomeration dwindles south-westward through the Piedmont until the Cotton Kingdom with its quite infrequent free Negro is reached. The upper and middle Ohio Valley had become the principal population region inland from the Atlantic Slope with south-western Pennsylvania, south-western Ohio, the Blue Grass Region, and southern Indiana standing out rather conspicuously. Both New Orleans and the rural Louisiana population had increased rapidly. It is probable that by then the long-distance migrations had begun in earnest, a point in evidence being the situation in the Ohio Valley. The presence of nearly 2000 free Negroes in eastern Tennessee suggests it as a sanctuary for fugitive slaves; the annexation of Florida has, however, added fewer free Negroes than might have been anticipated from the excitement generated in Georgia over slaves departing in that direction.

By 1850 the population was fully developed in the South, and nothing need be noted except the general scarcity of free Negroes in the mountains of West Virginia and the very small number present in Texas. The Texas situation appears to have been complex with negroid populations entering from Mexico and legal and illegal immigrants from the United States. For reasons unknown, the free Negro population may have shrunk greatly in the period just prior to American annexation, but we cannot rely on the available statistics 4. In the North, urban populations appeared at the major Lake ports, and a scattering in southern Michigan and a cluster along the Wabash can be detected. Many moderately large urban and rural nodes appear in south-western Ohio, which seems clearly to have become a favourite destination for the migrant. With the exceptions noted, the free Negro population of the Middle West was sparse and scattered. And, finally, several hundred free Negroes-mostly male-appear in central California.

Figure 4

The eve of the great civil conflict found the pattern of 1850 relatively unchanged. The Ohio and Indiana colonies had been reinforced, and southern Michigan emerges as a significant secondary area with a large settlement in Cass County. On the actual frontier of settlement the free Negro was virtually non-existent except in central California. There the number of Negroes in mining camps was rather large, and San Francisco contained a sizeable coloured population. The near-stagnation or actual retrogression of population in much of the Deep South may be noted.

The Canadian situation calls for brief comment 5. The figures for this country, whose coloured population came so largely from the United States, are mere guesses, but the best estimates agree on 30,00040,000 Negroes in the period just before 1850 when the majority were concentrated in Ontario with a significant number residing in the Maritime Provinces. By 1860 the coloured population of Canada may have increased by some 10,000, but after 1865 a large number of these Negroes drifted back to the United States.

The regional changes in population distribution can be better understood by consulting Fig. 5 where the percentage of the total free Negro population contained by each of eleven regions is shown for the years 1790-1860. The steady position maintained by the states along the Atlantic Slope, which held the principal concentrations, is plainly demonstrated; these were areas that never lost their supremacy, and they stand in sharp contrast to the great dwindling off in New England. Of the inland areas, only the Ohio Valley matured into a major population region. A study of Figs. I and 6 brings out the areal changes in the relative importance of the free Negro. There were several areas in the United States which experienced marked increases between I790 and I86o, but the most important was again the region along and near the coast from Massachusetts to North Carolina. In contrast to this hearth area, the free Negro was generally losing out to a more vigorous native white and immigrant influx in the newly settled regions. While he was slipping in numerical significance in the nation as a whole after the first quarter of the nineteenth century, the free Negro continued to advance in the one area where he had always been present in force. In the i85o-6o decade there appear to have been few decided areal changes in the relative significance of free Negroes.

Figure 5

The semi-logarithmic plotting of population growth by region for both total and free Negro population enables us by its representation of relative rates of increase to note the closely parallel growth patterns in the free Negro and the aggregate populations in Kentucky and Tennessee and in the Ohio Valley and the distinctly disparate rates in the North Central States and the Deep South. Other regions demonstrate an intermediate degree of correlation. Interestingly enough, the rate of free Negro population growth does not equal that of total population growth in even a single region after I840, although it had been greater in the period 1790-I8Io in the Upper South and was roughly parallel with the increase in aggregate population there for the next 30 years.

As regards the percentage in the total Negro community in the south in I790 and 1850, except for a not too prominent heightening of values in the upper reaches of Chesapeake Bay, the I790 pattern is not particularly meaningful and possibly indicates a stage of rapid but confused evolution; but by 1850, when the free Negro population had in many respects reached its climactic development in the south, the north-eastern quadrant of slave territory shows the greatest values for the free Negro with a crudely defined increase toward the north. The Allegheny Plateau, the Ozarks, and northern Missouri appear as virtually non-negroid except for a few isolated cases which are much more conspicuous on the map than their real importance would warrant. Except in the Mississippi Delta, portions of the Gulf Coast, and northern Florida, the free Negro is a rarity in the Deep South, and the influence of French and Spanish settlement immediately suggests itself in the exceptional areas.

The explanations for the distributional aspects of the free Negro hinge largely on regional differentiations in manumission rates and the direction of significant migrations. Large numbers of slaves obtained their freedom by means of manumission in Delaware, Maryland, Virginia and portions of North Carolina, but it was exceptional for a slave to be liberated voluntarily in the newer areas of the south. An economic explanation is obvious: the former regions were those which experienced a protracted agricultural depression after the American Revolution, and the period of most rapid manumission coincided with the time of maximum economic distress. Slaves were a liability on the worn-out farms of the Upper South, but where they were in great demand, as in Arkansas, Mississippi, or western Tennessee, the prospect for manumission was dim, and the same can be said of Louisiana after I840. Legal restrictions against the free Negro certainly played a strong part in determining population patterns in the South and account, for example, for the trend toward disappearance in Arkansas and Mississippi. The concentrations in the Blue Grass Region and in the Nashville Basin may have been largely results of the economic attractions of these superior areas for free migrants. Historical factors can be invoked to explain the situation in Louisiana where the French and Spanish socio-economic milieu had created conditions favourable for the growth of free Negro population. There was a universal and quite natural tendency to gravitate toward the large cities which was especially strong in the north-east. In the rural areas of the Middle West we can observe rather random bunchings of free Negroes which were largely the work of colonization or resettlement organizations. The western Pennsylvania agglomeration may be the combined result of a strong demand for mine and factory labour and the termination of several escape routes from the south. This latter factor was even more operative in the case of Indiana and especially Ohio, for a good many of the Underground Railroad lines led into this narrowest corridor between the south and Canada. The accessibility of the state and the strength of the abolition movement there were major factors in making Ohio one of the chief concentration points for fugitive slaves as well as for legal migrants.


In the census of1860 and in earlier enumerations, data on internal migration are crude or entirely lacking. This situation is aggravated in the case of the Negro, so that we must rely largely on indirect sources of evidence for any understanding of migration patterns. In Fig. 7, which depicts the sources of the free Negro population of eight of the largest American cities in 1860, the only available body of published data on migration has been presented graphically. The patterns for New York and Philadelphia are surprisingly similar to present-day movements of Negroes to these cities, and adding the Boston data, we have a strongly developed south-to-north axis of migration, much of which must have been by sea. The immigrants from Europe are a curious and puzzling item; but the small movement from the West Indies is a foreshadowing of things to come. The Cincinnati and Chicago data best express two major movements which have combined to form that strongly north-westward vector which has characterized Negro migration both before and after the Civil War: a sizeable stream of migrants flowing from the south converging with one moving westward from the north-east. In distinct contrast to the northern cities, most of the population in Baltimore and New Orleans were born in their respective states. Evidently, migrants from other slave states by-passed these metropolises, preferring to settle instead in the north. It may be supposed, nevertheless, that Baltimore and, to a lesser degree, New Orleans absorbed many migrants from nearby districts within the state.

Figure 6

One may argue deductively that migration, both legal and illegal, must have been especially vigorous in those areas where slave territory adjoined free soil. This notion may be tested and some index to the volume of Negro migration acquired by comparing changes in the coloured population in the border counties of border states with those that occurred in the remaining counties 6. This has been done and the results tabulated in Table 2.

Although there is little factual proof at the present time, it is plausible to suppose that the great majority of the 4553 free Negroes making up the deficiency in 1860 were migrants who moved northward, but the situation for the slaves must have been more complicated. In their case not only was there greater temptation and opportunity for escape, but the slave owners may have been more inclined, because of this hazard, to sell or transfer their chattels to less marginal locations or else to manumit them. Nevertheless. this somewhat fictive parcel of missing slaves must have been predominantly composed of migrants to the North, and if the greater part of 19,050 Negroes moved from this relatively limited area in only io years, the figure for all of the South must have been quite large in the full course of time.

Table 2. Discrepancies in Negro population changes in border counties of border states as compared with other counties, 1850-60

The sex ratio of free Negroes has been studied by states with the aim of discovering clues as to the nature and direction of free Negro migration. The deficiency of males that is particularly acute in the Slave States may be ascribed to a manumission differential favouring females and to the probability that a somewhat greater number of males than females participated in long-distance migrations, and especially in the Underground Railroad, a hypothesis which the strong male components in the states of the Old North-west tend to support. The very great preponderance of males on the west coast is chiefly a manifestation of frontier demography. The nearly normal ratios in the relatively undisturbed areas of northern New England may represent the end-result of a long period without manumissions and of low mobility, while the preponderance of females in Pennsylvania and New York is partially explained by the strong attraction of their principal metropolises for female migrants. The free Negro populations in Minnesota, Kansas, Arkansas and Texas were probably too small for their sex ratios to be statistically significant.

Figure 7

In Table 3 we have listed the sex ratios for all cities containing 500 or more free Negro inhabitants in x86o. In the great majority of cases the female element is much more strongly represented than the male even when compared with the state figure for free Negroes, and the fact of this quite uneven parcelling of the sexes in cities must have been apparent to the people involved. The existence of a high ratio in Chicago, Buffalo, Troy, Cleveland and Pittsburgh, all of which were important ports or stopping places on the routes to Canada, lends credence to the contention that the physical hazards of the Underground Railroad favoured the participation of males. The total picture, then, conforms closely with the accepted theory that males predominate in long range migrations and females in the shorter migratory movements.

Table 3. Sex ratio in urban free Negro populations of 500 or more, 1860
Table 3. Sex ratio in urban free Negro populations of 500 or more, 1860

A final but rather difficult method for studying migration is by means of age composition. The multiplicity of factors which influence the relative importance of the various age categories severely limits the usefulness of this class of information. Manumission differentials, differential fertility and mortality, differential migrations and the various peculiarities of the large base population to which the group is referred make for a complex interplay of causes. With these reservations in mind, it may be stated that a deficiency in any age group, except the youngest, indicates emigration, and an excess immigration. The situation for Ohio is rather too ambiguous to permit interpretation unless we recall that the state had been receiving as well as sending out large numbers of migrants for half a century and would have achieved an age distribution pattern peculiar to it; but the data for Michigan clearly indicate the in-migration of young adults, and especially males, who are the parents of a rather large number of children. The newness of the Michigan population is suggested by the deficits in the older ages. In the South, Virginia, Delaware and North Carolina conform nicely with the powerful certainty that these were major source areas for migrants. Except in the case of Delaware, the movement was predominantly male. The age composition of the northern cities is, as expected, quite abnormal, and a recent acceleration in the influx of young adults may be indicated by the general scarcity of the aged. In the southern cities a somewhat less abnormal situation with a greater proportion of children and the aged might imply that these cities received fewer migrants than did the northern metropolises. From the foregoing facts and analyses we can infer very little of value concerning the absolute volume of either the legitimate migration of free Negroes or the Underground Railroad except that a good many tens of thousands of individuals were involved. Nevertheless, such facts might be worked out by means of more refined statistical techniques. We have in this paper traced the origins of the free Negro population of the United States, noted its characteristics, chronicled its distribution, and attempted explanations for these spatial patterns. We have closed with a brief and highly imperfect essay on the migrations of this group, and as a parting suggestion it can be hoped that closer investigation of an obscure record may yet clarify the outlines of a significant shift in population and one of the more poorly charted areas of American historical geography.

1. The best general summation of our knowledge concerning the free Negro and an excellent bibliography are presented in John Hope Franklin's From Slavery to Freedom. A History of American Negroes (New York, 1948).
2. Wilbur Zelinsky, “The historical geography of the Negro population of Latin America”, Journal of Negro History, April 1949.
3. Melville J. Herskovits in his The American Negro has presented the anthropometric evidence for extensive miscegenation in the ante-bellum Negro population. This is perhaps the soundest approach to a question that is otherwise encumbered with the personal prejudices of the student.
4. Cf. H. Schoen, “The Free Negro in the Republic of Texas”, Southwestern Historical Quarterly, Vol. XL, April I926 and succeeding issues.
5. Cf. Ida C. Greaves, The Negro in Canada, McGill University Economic Studies, no. 16, Montreal, 1930.
6. I am indebted to Siebert (The Underground Railroad, p. 379) for this idea, and have attempted to carry it somewhat further than he was inclined to.