Chapter 3 provided the analytical framework that explained the interactions among the growth of population, economic growth, and the spread of diseases. Here we present historical and biomedical evidence that places the model in the context of nineteenth-century America. The available and pertinent data are on (1) the growth of population, (2) urbanization and population density, (3) transport developments associated with the demographic changes, and (4) the prevalence of infectious diseases across states and time. Some of the data are well known and widely accepted; however, including census mortality data, evidence on diseases is less reliable than that for population and transport. The evidence that does exist is consistent with the conceptual underpinnings that combine an increase in market size and specialization with an increase in infectious diseases.
The emphasis is not in the novelty of our story of the impact of diseases during the nineteenth century, but in the similarity of the nineteenth century with past centuries. Like the nineteenth century, all economic expansions of civilizations in the past were eventually wracked by deteriorating disease environments. The nineteenth century was not unusual because diseases became more widespread; what was unusual in the nineteenth century was the incubation and initiation of developments during the latter half of the century that were to break the age-old linkage between population growth, economic expansion, and a deteriorating disease environment. The story of how and why that linkage broke is an important one, but it is not our story. In this chapter, we provide evidence on how the economic expansion of nineteenth-century America is consistent with our population and disease model of economic growth and is similar to the past as it facilitated the spread of pathogens throughout most of the century.
In 1800, the population of the United States was approximately 5.3 million people, and the nation’s borders encompassed approximately 864 thousand square miles. Most of the population was clustered along the Atlantic coast with the center of population of the United States located just west of Baltimore (Atack and Passel 1994, p. 244). No region of the United States had a sufficiently dense population to warrant the capital investment required by an extensive canal system. Competition from natural waterways (the Hudson, Connecticut, Potomac, and other rivers) and the lack of concentrated economic activity inhibited the development of canals. The first canal in the United States was in operation before 1800, but the canal (in South Hadley, Massachusetts, built in 1796) was small and was located in a relatively densely settled area that connected populated areas along and to the Connecticut River. With the passage of time a system of inland transportation improvements became feasible with increased population and economic output, the settlement of inland territories, and the concomitant growth of agricultural production.
Canals had been built in antiquity, and canal locks were being built by at least the end of the fifteenth century. Canals were not built earlier in the United States because the size of the American market was inadequate to make the absolutely large financial investments that canals required economically viable. In general, highly specialized investments that require a substantial investment in fixed capital are not economic in markets that are small (relative to the financial commitment) and fragmented. This is why specialized investments in wharves, warehouses, and port facilities in British North America were delayed until the eighteenth century rather than earlier. Walton (1967) and North (1968) attribute most of the growth in productivity of ocean going transport in the eighteenth century to the development of specialized facilities (including protection from piracy) that were noneconomic earlier because the market was too small.
Population growth averaged close to 3 percent per year throughout much of the nineteenth century; at 3 percent, population doubles every 23 years. Table 7.1 shows the growth of the American population during the nineteenth century. The population grew from 5.31 million people in 1800 to 76 million in 1900. The peopling of the interior was even more rapid. The population center of the United States moved from just west of Baltimore in 1800, to nearly 100 miles further west in Virginia in 1820. By 1860, the population center of the United States was in Ohio; in 1900, it was in the middle of Indiana. In the 100 years of the nineteenth century, the geographic center of the American population moved more than 500 miles west.1
These shifts in the population center are for the entire United States; the movement of slaves mirrored that of the free population. With the expansion of the cotton economy and antebellum cotton plantations into the New South, the slave population and the pathogens that slavery harbored followed. In the five states of the New South (Alabama, Arkansas, Florida, Louisiana, and Mississippi), the slave population increased from 145,374 in 1820 to 1,376,297 in 1860, an increase of 847 percent, or approximately 5.8 percent per year. The free population of these states increased from 225,751 in 1820 to 1,663,086 in 1860, an increase of 637 percent, or approximately 5.1 percent per year (North 1961, table 9, p. 129). Data on the interregional movement of slaves between 1790 and 1860 also indicate a dramatic increase in the regional movement of slaves within the South. The changes captured in the data are impressive: 17,000 slaves were shipped interregionally in the 1790s; 31,000 slaves were shipped in the 1800s; 101,000 in the 1810s, 121,000 in the 1820s; 223,000 in the 1830s; 149,000 in the 1840s; and 193,000 in the 1850s (Fogel and Engerman 1974a, p. 46). These movements of people (enslaved and free) contributed to the spread of pathogens and facilitated the nationwide spread of diseases that previously had been local and regional.
The population density of the nation increased rapidly during the nineteenth century. The data on population density in the United States for 1790 to 1900 also are reported in table 7.1. The data indicate that population density in the nation increased by 21 percent between 1800 and 1830, 43 percent between 1830 and 1860, and another 141.5 percent by the turn of the century in 1900. The geographical boundaries of the United States approximately tripled during this era, making the increased density even more impressive. The growth in density varied widely across states though; the population per square mile in Massachusetts in 1800 was 52.69; in 1860, it was 153.14; and in 1900, it was 348.97. In Ohio, the corresponding population densities for the same years are 1.05, 57.43, and 102.05; in Pennsylvania, the corresponding densities are 13.44, 64.82, and 140.57; in Virginia, the corresponding densities are 20.07, 30.29, and 46.05.2
Along with increased density came a more than proportional increase in the urban population. In 1800, the percentage of the total population classified as urban was 6.1 percent; in 1830, it was 8.8 percent; in 1860, it was 19.8; and in 1900, 39.8 percent of the population was classified as urban. The rapid growth in the number of urban places and cities can be seen from the data in table 7.2, which reports the number of urban places for various size categories, total number of urban places, and the percentage of the population classified urban for 1800 to 1900. The total number of urban places grew 173 percent from 1800 to 1830, 336 percent from 1830 to 1860, and 344 percent from 1860 to 1900. For larger urban places, the growth in the 50,000 to 100,000 size category was 200 percent from 1800 to 1830, 133 percent from 1830 to 1860, and 471 percent from 1860 to 1900. The increases in population density and urban places and the growth in the size of cites assisted the spread of diseases; these demographic changes allowed pathogens to survive in places where they would not have survived in earlier centuries when humanity peopled the land less densely.
Increasing population and output increased the demand for transportation services. This led to investments in specialized transport services that both lowered the average cost of shipping a given volume of freight and increased the speed of transportation. These investments included the historically familiar canals, steamboats, and roads and the less familiar wharves, harbors, navigation charts, and lighthouses. Within the United States, investments in roads, canals, coastal packet ships, steamboats, and tugboats led to substantial declines in the costs of transport, and concomitant increases in the volume and frequency of interregional traffic. As the lower Mississippi became settled with slave plantations, they generated an increased demand for shipping services. The subsequent increased specialization in the transport sector led to falling costs. These transport developments were so pervasive and rapid that freight prices on steamboats on the Mississippi for the Louisville–New Orleans route declined extraordinarily quickly. In 1830, an index of freight prices on that route was just 25 percent of its 1815 level; and by 1860, it was just 11 percent of its 1815 level (Mak and Walton 1972, p. 639, app. table II). Table 7.3 presents data on average freight rates for each direction between New Orleans and Louisville in the years prior to the Civil War. Relative to pre-1820 rates, upstream freight rates by 1850 fell 95 percent; downstream freight rates by 1850 fell less, but they were still less than a third of their pre-1820 rates.
Data for freight transport costs in the United States circa 1815 and circa 1860 are presented in table 7.4. The costs of all forms of transport fell substantially between 1815 and 1860. Improvements in roads lowered freight costs in 1860 to about half their 1815 levels. Improvements in sailing ship design and harbor management made ocean shipping the cheapest form of transport; in 1860, the cost of ocean transport was 95 percent less than its costs circa 1815. Improvements in sailing ships on the Great Lakes and competition from steamships gave the Great Lakes the second lowest freight costs in 1860. If we proxy the missing data for shipping on the Great Lakes for 1815 with that of ocean shipping circa 1815 (1 cent per ton-mile), then the estimated decline in shipping costs on the Great Lakes from 1815 to 1860 was 90 percent. In a similar vein, investments in canals had a major impact on transport costs; Daniel Drake claimed that “[C]anals drove down the cost of shipping from about 20 cents per ton-mile in the 1810s and 1820s to 2 or 3 cents per ton-mile in the 1830s” (quoted in Atack and Passell 1994, p. 155).
Antebellum steamboat shipping had an enormous impact on lowering freight rates from 1811 to 1868. As a consequence, total tonnage on the western rivers increased from 400 tons per year in 1811 (at the beginning of the steamboat era) to 14,500 tons in just ten years (1821); by 1846 shipping tonnage had increased to over 100,000 tons (a 633 percent increase over 1821). Fourteen years later in 1860 the annual tonnage on the western rivers was 195,000 tons (an 83 percent increase over 1846). (Annual data on the number of steamboats and total tonnage (or capacity) on the western rivers are presented in table B.1 in appendix B.)
Throughout the antebellum years, accompanying declining transport costs and the increasing movement of goods were dramatic decreases in the times required to carry freight (and passengers). Travel times by inland water transport fell by more than 50 percent on some routes between 1815 and 1840. On the Mississippi, an upstream trip by steamboat from New Orleans to Louisville in 1815 to 1819 took 20 days; by 1840, it took 9.5 days; by 1860, it took just 6.5 days. While the declines in downstream travel times were not as dramatic, they were still sizable; from Louisville to New Orleans travel times for the corresponding years in order were 10 days, 6.6 days, and 5.2 days (Mak and Walton 1972, p. 630, table 2).
Indicative of the magnitude of the “less familiar” types of transport investments is the growth of federal expenditures on rivers and harbors; they grew substantially over the entire century, but not evenly. Beginning with $1,000 in 1822 to nearly $19 million in 1900, yearly federal expenditures fluctuated unevenly through the early 1860s. Much of the investment on transportation through the mid-nineteenth century came from sources other than the federal government (private sources and state and local governments). Still federal expenditures grew from 1822 until 1837 when they reached a pre-Civil War peak of $1,362,000. Subsequently, annual federal expenditures fluctuated with no obvious trend until the end of the Civil War. After 1866, federal expenditures on transportation began a long secular increase that continued to the end of the century. For the twenty years 1867 to 1886, expenditures averaged $6.4 million annually; they averaged $9.8 million annually for the entire post–Civil War period—1867 to 1900. (Annual federal government expenditures in the United States on rivers and harbors for the years 1822 to 1900 are presented in table B.2 in appendix B.)
Investments in transportation infrastructure predate any significant federal expenditure. The first major canal built in the United States, and the one that set off a canal building boom was the Erie Canal, was initiated by the state of New York. Started in 1817 (twenty-one years after completion of the first American canal in South Hadley), construction on the Erie Canal was completed in 1825. The Erie Canal provided low-cost bulk shipping to the nascent economies of the upper Midwest and the Great Lakes. The Erie was built in segments from east to west; segments were opened as soon as they were completed; these segments were immediately profitable. Well before its terminus on Lake Erie was reached, one million dollars per year in tolls were being collected. The entire multimillion-dollar cost of the canal was recouped within a few years of its completion (Atack and Passell 1994, p. 150). The immediate profitability of the canal is consistent with an increase in the size of the market leading to increased specialization; alternatively, the potential demand for transport services preexisted the Canal. The construction of the Erie began only after New York State west of the Hudson River Valley and, more generally, the Great Lakes region had experienced sufficient population growth and settlement to generate a large potential demand for transport services.
From 1837 to 1860, tonnage on the Erie increased 238 percent, from 667,151 tons to just over 2.25 million tons; it increased another 105 percent to just over 4.6 million tons by 1880. Tonnage data indicate the continued growth of shipping between the Great Lakes region and the port of New York. The Civil War era and the years immediately thereafter were enormously profitable for the Erie; the data indicate a massive diversion of trade during the War (and for a few postwar years) from the North–South river network to the East–West route of the Erie Canal. (Annual data on tonnage in the mid-nineteenth century for the Erie division of the New York state canals are presented in table B.3 in appendix B.)
Following the Erie’s early success, a host of other canal projects were initiated; the Pennsylvania Main Line, the Schuylkill, the Champlain, the Delaware and Raritan, the Chesapeake and Ohio, the Ohio and Erie, the Miami and Erie, the Wabash and Erie, and the Chicago–La Salle Canals (to name some of the major undertakings). These canals were completed during the three decades subsequent to the Erie’s completion. By 1860, the United States had 4,254 miles of completed canals, and the canal era had ended. Canal building stopped because competition from railroads did not allow canals to be profitable. (Only during the years impacted by the Civil War did canal prosperity revive only to disappear again with the return of normalcy.)
Developments in transportation allowed an extraordinary increase in the movement of goods across regions. An example is the development of the Ohio and Erie Canal; the canal’s first leg between Cleveland and Akron opened in 1827. The entire canal was completed in 1832, a total distance of 308 miles between Cleveland on Lake Erie and Portsmouth, Ohio on the Ohio River. Within one year of its completion, wheat shipments through Cleveland had increased from 1,000 bushels to 250,000 bushels; seven years later in 1840, wheat shipments through Cleveland were 2.2 million bushels.
The first miles of railroad track were being laid just as canal building was in full bloom in the nation. In 1830, there were twenty-three total miles of railroad track in the United States; in 1840, there were 2,818 miles. The railroad industry expanded rapidly in the 1840s and by 1850 total mileage had increased to 9,021 miles, more than twice the total mileage that canals had in 1860. There were over 30,000 miles of railroad track in 1860, a 239 percent increase from that of 1850. The growth of the nation’s railroad network continued during the post–Civil War decades of the nineteenth century; by 1890, there were 166,703 miles of track, 444 percent more than in 1860. (Annual data on railroad track during the nineteenth century are presented in table B.4 in appendix B.)
In their infancy railroads complemented rather than substituted for canals; in the 1830s, shipping by rail was substantially more expensive than canals for most purposes. When railroads were first constructed, it was where canals were very costly to build or maintain—over hills and mountains, and between ports and cities. Railroads were cheaper to build than canals, but they still had substantial fixed costs; their variable costs, which were lower than other forms of overland transport, were higher than those of canals. This meant that passenger fares and freight rates on the early railroads were lower than other forms of overland transport but not lower than on canals. Consequently, rail mileage in the 1830s and 1840s was concentrated in the densely populated areas of the Northeast where the volume of traffic was greatest and passenger tolls were a significant portion of the income of railroads. (Passenger traffic is, and was, willing to pay a great deal for speed; a 200-mile rail trip took about half a day in 1840; the same trip by canal boat would take over three days.) In the 1840s, more than half the total railroad mileage in the country was concentrated in the Northeast. Continuing improvements in engine efficiency, rails and track, and other innovations gradually reduced the costs of rail traffic. As a result, railroad track mileage and traffic increased substantially. In 1860, developments in rail transport had given railroads at least three distinct advantages over inland water freight: (1) railroads could be constructed in places where canals were impracticable, (2) railroads could be operated year round, and (3) railroads were faster than water transport.
Railroad travel times in 1860 were small fractions of those in 1830. The percentage declines in the time it took to travel long distances (primarily along East–West routes) were greater than the declines for shorter distances. Travel times circa 1860 were about 1/7 to 1/10 of those in 1830, which in turn were about half of those in 1800. In 1857, eastern Ohio was one day’s travel from New York City; it had been a seven-day journey in 1830. The travel time to New York City from much of Illinois was two days in 1857 compared to journeys of three weeks in 1830. In the densely populated urban areas of the Northeast, travel times between cities were typically a matter of hours.3 Increased speed allowed railroads to capture nearly all the higher valued freight from water transport; the speed and comfort of railroads redirected passenger traffic from canals and steamboats to the rails as well.
In trans-Atlantic shipping, declining costs in both time and money affected the immigration of Europeans to the United States. Steerage fares from Europe to the United States fell by almost 70 percent between the 1810s and the 1830s. By the late 1820s, packet ships had regularly scheduled crossings of the Atlantic, and fares were 50 percent less than the fares a decade earlier (the 1810s). These developments aided the movement of peoples and the pathogens that preyed upon them. Two significant changes in immigration to America were (1) a substantial increase in the numbers of people coming to the United States and (2) major changes in the primary countries of origin. Changing origins implies that different (to Americans) pathogens entered the United States in the nineteenth century. Early in the nineteenth century, the number of immigrants was relatively small; in 1820, total foreign immigration was just 8,385 people; in 1830, it was 23,322. By 1840, immigration was 84,006; from 1846 to 1860, immigration averaged over a quarter million people per year (Carter, Gartner, Haines, et al. 2006, vol. 1, series Ad1–2, pp. 1–541). The entire population in 1845 was just over 20 million; consequently, the foreign born added about 1 percent per year to the American population for each year from 1846 to 1860. Immigration reached its pre–Civil War high in 1854 when 427,833 immigrants came. The early Civil War years reduced the flow of immigrants considerably, but by 1863 it was recovering. Immigration averaged more than 365,000 people annually from 1863 to 1900. Nineteenth-century immigration peaked in the 1880s when an average of nearly 525,000 immigrants came each year of the decade, reaching a single year high in 1882 of 788,992. (Annual immigration and annual immigration for the three major countries of origin are presented in table B.5 in appendix B.)
Nineteenth-century developments in transportation, the increase in the volume of freight, and the movements and numbers of people affected the transmission of diseases. In prior centuries, diseases were primarily local and regional. In the nineteenth century, the growth of transport networks, the reduction in transport costs, and the reduction in journey times led to an integration of disease pools: locally, regionally, nationally, and, in some instances, internationally. The diseases that were identified and spread included (among others): cholera, tuberculosis, dysentery, malaria, hookworm, influenza, gastrointestinal diseases, measles, yellow fever, mumps, and smallpox. These diseases had existed before the transport changes of the nineteenth century, but they were generally localized and regional. A widespread, low-cost, and relatively rapid, transport system integrated goods, peoples, and, importantly for our story, diseases. The increasing density and the growth of cities and urbanization further integrated disease pools and facilitated their propagation. Increased virulence accompanied disease strains when “new” diseases or their variants were introduced to areas where they had been absent.
Table 7.5 presents data on malaria morbidity rates for American soldiers in the 1830s. At the time, malaria was known under a variety of names, such as “autumnal, bilious, intermittent, remittent, congestive, miasmatic, malarial, marsh, malignant, chill-fever, ague, fever and ague, dumb ague, and lastly the Fever” (Drake [1850, 1854] 1964, p. 703). The data in table 7.5 are for military posts of the United States Army in coastal areas and the interior of North America; the data are reported by latitude and name of the post in the 1830s. Although the latitudinal location is known for all posts and the location of many is obvious, the exact location of some posts has to be determined from the historical record and maps.
The morbidity rates in table 7.5 indicate that malaria had spread from coastal areas in the southern United States to northern areas along the Mississippi River system. While the prevalence of the disease generally declined the further north a military post was located, other geographical aspects of a post’s location also were important to the survival and propagation of the malaria. For example, Fort Towson (in present-day Oklahoma) was more northern than both Fort Mitchell (Alabama) and Fort Jesup (Louisiana), yet it had a malaria morbidity rate that was from four to five and a half times higher than the other two forts. Fort Towson had a substantially higher incidence of malaria in spite of its more northerly location because it was located in a low-lying area a few hundred yards from a small standing lake right in the heart of the Red River basin. (The Red River is a major tributary of the Mississippi.) Fort Jesup was located in a part of northern Louisiana about 120 miles west of the Mississippi River; although it has relatively mild winters, it receives measurable snowfall once every five to ten years (and killing frosts more frequently). Fort Mitchell was located about 200 miles north of the Gulf coast near present-day Auburn, Alabama, in a partially hilly area at the edge of the Piedmont plateau. So it too had geography less amenable for the propagation of mosquitoes and the spread of malaria than Fort Towson.
If we allow the incapacitation rates in table 7.5 to approximate the general prevalence of malaria, then the data indicate how pervasive malaria was in the antebellum era. In some areas, the incapacitation rates for presumably relatively healthy young men (they were in the military) exceeded 100 percent (more than 1,000 cases of incapacitation due to malaria per 1,000 troops per year). At these posts, the troops averaged more than one malarial spell per year per soldier. The morbidity rates indicate that the soldiers experienced malarial attacks that removed them from the duty roster some time during the year. The data do not indicate the length of incapacitation.
United States Census data on the number of deaths by specific causes in each state for the years 1850 to 1900 provide further evidence of the spread of infectious diseases during the nineteenth century. From these data, we calculated cause-specific mortality rates for the entire population for nearly two dozen major infectious diseases for the 1850 to 1900 census years for each state for which data exist. We also calculated mortality rates for deaths from all causes for (1) the entire population, (2) four age cohorts—adults (20 years and older), children/adolescents (5–19 years old), young children (1–4 years old), and infants (under 1 year old), (3) two ethnicity cohorts (blacks and whites), and (4) two nativity cohorts (native born and foreign born) for each census year for each state for which data exist. The cause-specific mortality rates for the major infectious diseases are presented in tables C.1 through C.11 in appendix C. The mortality rates for all causes of death for the entire population and the eight different cohorts are presented in tables C.12 through C.16 in appendix C. Appendix C also includes a discussion of issues involved in the calculation of the mortality rates for several of the diseases, an examination of the overall accuracy and reliability of the census mortality data, and a listing of the data sources.
The cause-specific mortality rates indicate that many of the infectious diseases had spread from the locations where they were initially concentrated to other regions of the country. Expanding on the evidence on malaria morbidity presented in table 7.5, the malaria mortality rates indicate that malaria had clearly spread far beyond the South, and by 1850 it was embedded in the Midwest and much of the Northeast. But malaria had reached its peak prevalence in the Northeast by the mid-nineteenth century, from then until the end of the century malaria mortality rates in most northeastern states declined steadily. By 1900, malaria mortality rates were small fractions of their 1850/1860 rates in many states in the Northeast. Similarly, malaria declined in midwestern states during the second half of the nineteenth century; nevertheless, in several midwestern states malaria remained a nontrivial disease until circa 1880/1890. In contrast, malaria in the South remained a relatively major cause of death. Malaria mortality rates in the southern states were several multiples larger than the rates in rest of the country in 1900 (see table C.4).
Deaths from intestinal parasites/worms were concentrated in the Deep South (see tables C.5 and C.9). The data indicate that these diseases also afflicted the border states, parts of the Midwest, and at least two northeastern states; albeit the mortality rates indicate deaths in northern states were small fractions of those in the Deep South. For other diseases, the data indicate their epidemic nature; for example, mortality rates for cholera (all variants) in about two dozen states decreased from 1850 to 1860 to become a small fraction of their 1850 rates (see table C.1). Yellow fever, while not unknown outside of the South, was generally concentrated in southern cities, as indicated by the very high yellow fever mortality rates in Louisiana and South Carolina in 1850, in Texas and Louisiana in 1860, in Florida in 1870, and in Tennessee and Louisiana in 1880, also indicating the epidemic nature of the disease (see table C.9). The data also indicate that several of the infectious diseases appear to have declined throughout the nation, among these were croup (table C.2), dysentery (table C.3), scarlet fever (table C.6), and smallpox (table C.7). Conversely, other diseases appear to have increased, among these were diphtheria (table C.3), consumption/tuberculosis (table C.1), and pneumonia (table C.6).
We also calculated a series of regional mortality rates for the regional populations, and for the eight different cohorts for three regions of the United States for all causes of death and for four major groups of infectious diseases for the 1850 to 1900 census years. These regional rates aid in understanding the differences in the mortality rates for the regional populations and for the various cohorts across states and time for all causes of death and for the nearly two dozen infectious diseases that were examined. The regional mortality rates were calculated for three regions of the eastern third of the United States: (1) the Northeast (Connecticut, Maine, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, and Vermont), (2) the East North Central (Ohio, Indiana, Illinois, Michigan, and Wisconsin), and (3) the South (Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Virginia, and West Virginia). The four disease groups for which we calculated mortality rates are (1) upper respiratory diseases (consumption/tuberculosis, croup, influenza, pneumonia, “winter fever,” and whooping cough), (2) warm-weather fevers (bilious, congestive, intermittent, remittent, malarial, typho-malarial, typhus, and yellow fevers as well as fever, not specified), (3) intestinal parasites/worms (dirt eating, parasites, parasitic diseases, and worms), and (4) gastrointestinal tract diseases (cholera, cholera infantum, cholera morbus, diarrheal diseases, diphtheria, dysentery, and typhoid/enteric fever).4
The regional mortality rates for all causes of deaths indicate several interesting patterns. The mortality rates for the regional populations shown in figure 7.1 indicate that overall the Northeast was the least healthy region with the highest regional death rates. Northeastern death rates increased until at least 1890, ending the century substantially higher than their 1850 level. The data for the other two regions indicate that both regions ended the century with death rates essentially at their 1850 levels. Nearly all of the cohort mortality rates, except for foreign-born individuals, show the northeastern states with the highest mortality. Figures 7.2 and 7.3 depict the regional mortality rates for infants (under 1 year old) and young children (1 to 4 years old), respectively; the figures show that the Northeast had the worst environment for these cohorts. The highest mortality rates in the Northeast are consistent with our story of infectious diseases being a product of economic growth, increasing population, density, urbanization, large cities, and developing transportation networks.5
Infant death rates for all causes in all three regions were much higher than the regional rates for other age cohorts, and infant death rates do not appear to have fallen before 1890. The mortality rates for young children were second highest by far; they rose slightly in the Northeast and South, and fell slightly in the East North Central, from 1850 to 1880; they declined substantially from 1880 to 1900 in all three regions (see figures 7.2 and 7.3, respectively). The children/adolescents (5 to 19 years old) cohort had the lowest mortality rates. The foreign-born mortality rates generally are higher than the native-born rates, but the census mortality data categorized by nativity are the most problematic of all the population cohorts. In two of the censuses, deaths by nativity were reported only for whites (native-born and foreign-born whites); in another census, deaths by nativity were reported for whites and free “colored” combined (native-born and foreign-born whites and “free colored”); in another census, deaths by nativity were reported for the entire population (native born and foreign born). In the other two censuses, no mortality data were reported on the basis of nativity (see appendix C for more details).
The white and black mortality rates for all causes in the three regions are shown in figures 7.4 and 7.5, respectively.6 They indicate that during the 1850 to 1900 period blacks had higher mortality rates than whites, and black mortality rates were closest to white rates in 1850. Except for 1850, black death rates were the highest in the Northeast and lowest in the South, and black mortality rates exceeded white rates by the most in the Northeast and the least in the South. Except for 1850, black and white death rates were much more similar in the South than the other regions (compare figures 7.4 and 7.5). These findings are consistent with both the history of nineteenth-century America and our disease story for the second half of the century. Higher overall mortality for blacks during this period is consistent with inferior economic and living conditions for blacks in the industrial Northeast and the postbellum South (two-thirds of the census years examined come after the Civil War). In the northeastern states, the high black mortality rate and the large gap between the black and white mortality rates is consistent with crowded living quarters (density), urbanization, poor living conditions, and a greater relative susceptibility to “cold-weather” diseases. The similarity of the mortality rates for blacks and whites in the southern states is consistent with a combination of lower susceptibly of blacks to “warm-weather” southern diseases (and the increased susceptibility of whites), the increasing spread of diseases during the nineteenth century, and deteriorating social conditions for blacks in the postbellum South.
The regional mortality rates for the four disease groups indicate important patterns as well. For upper respiratory diseases, figure 7.6 shows the Northeast had the highest mortality rates for the regional populations, while the South and East North Central regions are generally similar. Infants (under 1 year) had the highest susceptibility to upper respiratory diseases with mortality rates triple those of young children (1 to 4 years), who had the second highest mortality rates. Similar to the white and black mortality rates for all causes of deaths, white mortality rates due to upper respiratory diseases, shown in figure 7.7, were much lower than those of blacks, depicted in figure 7.8, in the Northeast and East North Central regions while in the southern states blacks and whites had very similar upper respiratory mortality rates.
Figure 7.9 shows the mortality rates for deaths from intestinal parasites/worms for the regional populations; death rates in the South were several multiples greater than those of the other two regions. While there were some deaths from intestinal parasites/worms in both the Northeast and East North Central regions, they were nearly nonexistent by 1870 in the Northeast. Moreover, all three regions had declining trends from 1850 to 1900. In all three regions, the infant (and young children) mortality rates for intestinal parasites/worms were about forty to fifty times greater than those of adults (see the infant and adult mortality rates in figures 7.10 and 7.11, respectively) and ten to twelve times those for the children/adolescents cohort (not shown). The Census reported on deaths from intestinal parasites/worms by ethnicity (black and white) for only two of the six census years, making any conclusions about black and white mortality rates problematic.
The mortality rates for warm-weather fevers for the regional populations (see figure 7.12) indicate a common trend for 1850 to 1900 in the Northeast and East North Central states. Death rates from these fevers declined throughout the period (with the possible exception of 1880) in both regions until in 1900 they were no longer a major threat. The mortality rate for these fevers fell in the South so that by 1870 it was about a third of its 1850 rate; subsequently, there was a major resurgence in 1880 after which it returned to its (approximately) 1870 level and remained there. Similar to the findings for intestinal parasites/worms, warm-weather fever mortality rates were by far the highest in the southern states, especially for 1870 to 1900 when the southern death rates were two to about six times higher than those in the East North Central states and six to about eleven times higher than those in the northeastern states (see figure 7.12).
Among the age cohorts, infants and young children were by far the most susceptible to deaths from warm-weather fevers in all three regions. For the census years for which warm-weather fever deaths by ethnicity were reported (1850, 1870, 1890, 1900), the mortality rates for blacks (see figure 7.13) were greater than those for whites, except at the beginning of the period; in 1850, white mortality rates (see figure 7.14) were somewhat greater than black rates in all three regions. Because of the problematic nature of the census mortality data based on nativity (noted above), any conclusions about warm-weather fever mortality for native born and foreign born should be viewed cautiously. Nevertheless, the mortality rates for warm-weather fevers are highest for the foreign born in all three regions, except in 1850 in the South. But in the South in 1850, slave deaths were not reported on the basis of nativity and very few foreign-born individuals had settled there. As a result, the South’s native-born and foreign-born populations and deaths are not directly comparable to those in the other regions.
The mortality rates for gastrointestinal tract diseases for the regional populations, shown in figure 7.15, indicate the epidemic nature of the individual diseases that make up this group, with two major peaks in the death rates in 1850 and 1880 and major troughs in 1860/1870 and 1900. This pattern shows up as well in the mortality rates for young children, but less so for infants who were somewhat protected from waterborne diseases by relying mostly on mother’s milk. Gastrointestinal tract deaths do not appear to be noticeably concentrated in any specific region, except those for infants and young children that are noticeably greater in the Northeast. For whites and blacks, the mortality rates for gastrointestinal tract diseases are the most similar of all the disease groups across the four census years (1850, 1870, 1890, 1900) that have data for all three regions. The mortality rates by nativity are not comparable because too many of the individual gastrointestinal tract diseases were not reported for foreign born in the Census.7
The most susceptible elements of a population to infectious diseases are typically infants and young children. Infants who have just been weaned or are being weaned are particularly susceptible. For nursing infants, mother’s milk provides partial protection from infection for (at least) two reasons. The first is that breast milk directly provides antibodies to the infant (a breast-fed baby is less likely to be infected by diseases than is a non–breast-fed baby). The second reason is that breast-feeding eliminates the use of other foods that may harbor diseases (such as typhoid fever, dysentery, salmonella, and other diarrhea-causing diseases). It should be pointed out that while diarrhea is a symptom not a disease there are many infections of the gastrointestinal track that cause it. Many diarrheal diseases are frequently not identified, and even when they are, they are usually diagnosed as diarrhea rather than caused by a specific pathogen or pathogens. Infants and young children, because of their small size, are very susceptible to diseases that affect their gastrointestinal track. A human baby infected by an organism that causes diarrhea can dehydrate and die within a day. The enhanced susceptibility of newly weaned babies to diarrheal diseases is exacerbated by the infant's lack of exposure to disease-causing pathogens. In older children and adults, previous exposure to diseases primes the immune system with acquired antibodies that resist the same (or similar) invasive pathogens. An infant just weaned from mother’s milk has an immune system that lacks such prophylactic devices, so pathogens have a relatively easier time establishing themselves.
Consistent with the greater susceptibility of infants to infectious diseases, our state-level infant mortality rates in table C.14 in appendix C appear to indicate rising trends during the second half of the nineteenth century, at least for the eastern third of the United States. Other infant mortality rates for Baltimore, Maryland, the state of Massachusetts, and the Northeast region are presented in <!--table 7.6-->table 7.6; these rates show similar rising trends during the same period. In the second half of the nineteenth century, Baltimore was more populous and densely populated than the average city in the nation. Massachusetts was an industrial state where development and urbanization began earlier, and it was a destination for immigrants. Similarly, the Northeast region was more industrial and urbanized and had more immigrants than the United States as a whole. These are the very characteristics of locations that our disease story of economic growth suggests would have generated higher mortality rates especially among susceptible groups (that is, infants). Because the mortality rates we calculated for the Northeast are based on the census mortality schedules, which certainly underenumerated deaths for infants and for the earlier census years, they have some inherent limitations (see the discussion in appendix C). For example, it seems clear that the 1850 infant mortality rate for the Northeast in <!--table 7.6-->table 7.6 is a substantial underestimate compared to the other mortality rates shown in the table. Finally, while the infant mortality rates for the United States that are presented in <!--table 7.6-->table 7.6 do not indicate the same increasing trend as indicated by the other mortality rates in the table, they do indicate that the United States rate in 1880 was still over 98 percent of its 1850 rate.
The infant mortality rates in appendix C (table C.14) and table 7.6 are generally consistent with our disease story. The trends in many of the rates in table C.14 for the eastern part of the United States appear to indicate sizable increases, even if 1860 is used as the beginning date. These trends are more readily apparent in the Northeast region infant mortality rates (table 7.6 and figure 7.2). But given the certain underenumeration in 1850, the 1860 mortality rate is the better beginning date for examining mortality trends in the Northeast. The infant mortality rate in the Northeast in 1890 was over 41 percent higher than its 1860 rate, and the rate in 1900 was still 104 percent of its 1860 rate. (Moreover, as discussed in appendix C, infant deaths reported in the 1900 census no longer included stillbirths unlike the earlier censuses.) The Massachusetts data in table 7.6 show a 26 percent increase in infant mortality from midcentury to 1890 (Massachusetts infant mortality actually peaked in 1872 at 1,941 deaths per 10,000); in 1900, the Massachusetts infant mortality rate was still 18 percent above that of 1850; in 1910, it was still as high as that of 1850. The infant mortality rates for Baltimore show a 71.5 percent increase in the death rate from 1850 to 1890; in 1900, the mortality rate in Baltimore was still 45 percent higher than in 1850.8 (Our mortality rates for young children presented in table C.13 in appendix C do not show similar increases; they actually decline during the period but not by much, at least in the eastern part of the country. The 1890 mortality rate for young children in the Northeast region was still nearly 86 percent of its 1850 rate.)
As discussed in chapter 3, the Antebellum Puzzle is the term encompassing data that show adult heights declining in the birth cohorts of men who fought in the Civil War (born a couple or so decades prior to the war). The mean heights of age cohorts born after midcentury subsequently rose. The term “puzzle” is used because real living standards (as measured by real income per capita) were increasing during the period when adult heights were falling. In addressing this issue in chapter 3, we argued that the puzzle was explicable within the framework of a deteriorating disease environment. Here we add another “puzzle” that can be explained by the disease environment as well: Why were mortality rates of infants in the eastern part of the nation increasing as income per capita rose in the decades following the Civil War? Was there a “Postbellum Puzzle” to accompany the Antebellum Puzzle? The answer to these questions is embedded in the spread of infectious diseases. Infant mortality rates increased in the decades following the Civil War in those locations that were especially likely to have faced a deteriorating disease environment; infants are particularly susceptible to many infectious diseases.9 Similarly, in the antebellum United States, transport integration and revolution united disease pools even as they led to a more specialized and productive economy. Increasing urbanization facilitated the exchange of pathogens as well as economic exchanges. Cities of hundreds of thousands could only exist within a network of efficient, relatively rapid low-cost transportation that also materially assisted the spread of disease.
The increasing efficiency of transportation distorts nineteenth-century data that are used to measure living standards. In 1800, internal transport in the United States depended on natural waterways and animal power. By 1860, railroads, canals, and streetcars economized on the use of animal power in transport. Per unit of economic output there was literally less “horse” power in transportation in 1860 relative to 1800. Consequently, if the ratio of the stock of grain commodities to the population during the 1800 to 1860 period is examined, as some scholars do, the presumption of “holding other factors constant” is incorrect. The same amount of grain per capita or even a smaller amount may have meant an increase in grain consumption for the human population. Thus, it cannot be concluded from data on the stock of grains (or “food”) per capita alone whether a change in the amount of “food” led to a change in the biological standard of living.
Similarly, data on prices may give a misleading picture of the standard of living. The nineteenth-century developments in transportation reduced the wedge between what consumers paid for farm output and what farmers received for the output. Because of the transport improvements it is plausible that retail prices could have fallen while wholesale and/or farm gate prices rose during the first half of the nineteenth century. The reduction in the costs of transportation leads to higher prices to producers and lower prices to consumers. The wedge between the prices paid by demanders and the prices received by the producers narrowed due to the developments in the transport sector. This is the well-known consequence of any reduction in the costs of trade. As a result, it cannot be concluded from data on wholesale and/or farm gate prices alone whether a change in agricultural (wholesale and/or farm gate) prices led to a change in the biological standard of living.
In previous chapters, we argued that developments in transport are not exogenous; they are instead endogenous to the economy. Increases in specialization and output occur because of an increase in market size, this increase also increases the biological resources available for opportunistic pathogens. No ecological niche goes unoccupied for long. Diseases, morbidity, and increased infant mortality were some of the consequences; these affected measures of health and well-being throughout most of the nineteenth century. It was only with the increase in the knowledge of public health and biology, and improving water and sanitary practices that occurred relatively recently, did infectious parasitic diseases cease their slaughter of the innocents.