From the Editor: With some regularity we spotlight books in the tenBroek Library. Here is Librarian Ed Morman's description of another book in the collection:
Report on the Insane, Feeble-Minded, Deaf and Dumb, and Blind in the United States at the Eleventh Census: 1890, compiled by John Shaw Billings, M.D., Washington: Government Printing Office, 1895. 755 pages.
At the tenBroek Library we often get statistical questions that we cannot answer satisfactorily. Here are some hypothetical examples:
Why can’t we answer these questions? In some cases no one is collecting the needed data—and, if no one collects the data, no one can tabulate the statistics and publish them. In other cases the statistics are drawn from a sample and not adequately disaggregated; thus we may have a reasonable estimate of the number of disabled people in a given county but not necessarily broken down by type of disability. Beyond this we are often unable to compare data from different sources, because they define “blindness” differently.
The 2010 decennial census, which counted the entire population, did not include a question on disability. Instead, as part of the American Community Survey, the Census Bureau annually collects data on a relatively small number of households. This sample is scientifically chosen and therefore can produce statistically reliable approximations for the entire population of the U.S. It can also produce acceptable results for regions, states, counties, etc., but only for categories that include large numbers of people. If the numbers are too small, the Census Bureau cannot extrapolate from the sample, and it will not provide estimates it cannot stand behind.
In the 1960s the Department of Health, Education, and Welfare established a Model Reporting Area for Blindness. This consisted of thirteen states regarded as exemplary in their collection of data. As it turned out, there were serious problems with the data collected, especially underreporting of blindness among people of low socioeconomic status. There have since been other attempts to provide accurate estimates of the blind population, and a study in the 1980s documented the difficulties involved in a nationwide prevalence study of blindness.
Jernigan Institute staff members are currently working on a new statistics page for the NFB Website, in which we will attempt to provide the most current data that is deemed statistically reliable. We will announce its availability soon.
Meanwhile, let’s look back more than a hundred years, when the world was apparently simpler. The volume at hand was assembled for the Census Office by John Shaw Billings, deputy surgeon general of the U.S. Army.
Billings was an interesting character. He was trained as a doctor and was one of the leaders in bringing the germ theory of disease to the American medical profession. He was also a librarian, having founded the Army Medical Library, which has grown into the National Library of Medicine, the largest medical library in the world. And Billings was enough of a librarian that he later became the first director of the newly formed research division of the New York Public Library.
For those old enough to remember the punched cards we used to use with mainframe computers, Billings was involved with their use decades before they became ubiquitous as computer accessories. The cards are properly known as “Hollerith cards” and are named for Herman Hollerith, who worked under Billings at the Census Office. But that’s another story, unrelated to blindness.
So back to the volume at hand. First, we must recognize that Billings never made clear just what the census counted as blindness. But then, even today, is there a clear definition? There is, of course, the legal definition—but most Federationists prefer Kenneth Jernigan’s functional definition (see “A Definition of Blindness,” first published in the Blind American in November 1962, available online at <http://www.nfb.org/images/nfb/publications/fr/fr19/fr05si03.htm>). As for the Census Bureau, it currently chooses not to use the word “blind” at all and instead, asks respondents if they have “serious difficulty seeing even when wearing glasses.”
The 1890 volume tells us that the “total number of persons reported as blind in both eyes on June 1, 1890, was 50,568 . . . being 800 per 1,000,000 of the total population.” This was a decrease from 1880, and Billings suggested two possible causes for the decline over the previous ten years: “a real one, corresponding with the decrease which has occurred in England” or a statistical artifact, namely “less complete enumeration of this class of population than was made in 1880.” Even though Billings was not completely confident of the numbers collected in 1890, he tried to provide exact numbers for the blind in each state and even cross-tabulations by race, sex, age, and other variables. To explain differences from state to state, he suggested differences in “proportions of persons of advanced age” or “previous epidemics of eruptive fever in certain localities” or “perhaps . . . heredity.”
A special discussion on the congenitally blind appears to have been intended to confront the question of inherited blindness. We now know that not all congenital blindness is hereditary and that some hereditary causes of blindness do not manifest themselves until well into adulthood. Billings recognized this and tried to deal with inheritance by looking at how many blind people had blind relatives.
In addition to breakdowns by sex, age, race, and place of birth (if foreign born), Billings and his associates provided information on how many of the blind were in “schools or institutions for the blind.” Other tables and charts cover cause of blindness, marital status of the blind, and percentage who “can neither read nor write.” Those reported as congenitally blind were very unlikely to have ever married and were less likely to be literate.
Perhaps the most interesting table lists the occupations of blind men, broken down by age and race. Among the job categories for blind people in 1890s America were artists, bakers, bookbinders, carpenters, cigar makers, editors, engravers, fishermen, lawyers, laborers, factory operatives, miners, physicians, soldiers, and teachers.
After reviewing the contents of this volume, the question remains: why did the federal government, in a pre-computer age, go to such trouble to develop statistics on blindness (not to mention the other categories of disability discussed in this report)? Even with Hollerith cards and machines to tabulate them, this was a massive undertaking. Historians have suggested some answers to this question, and readers of the Braille Monitor might take a minute to consider whether they agree with the explanation provided below.
In 1890 the industrial revolution was in full swing in the United States. Railroads, mines, and factories—with their attendant dangers to occupational health and safety—were spreading around the country. America needed a growing workforce, and government and business leaders wanted to know what factors might contribute to lowering productivity. By breaking down cause of blindness to cover a range of industrial accidents, this report helped feed the growing Progressive Movement’s efforts to improve workplace safety. Counting the blind also enabled reformers, in and out of government, to estimate the cost of resources needed for education and rehabilitation. Even in those days, well before the founding of the NFB, some positive efforts were being made, and even those who encouraged dependency were often well-meaning.
A darker side of progressive reform was eugenics, the attempt to weed out “defective people” by preventing them from having children. Hence the discussion of hereditary and congenital blindness. Americans at the end of the nineteenth century were living in a new world in which science and technology played an augmented role. Reformers were attempting to make best use of the new tools, and sometimes the tools turned out to be dangerous.
This book is available in digital form on the Internet Archive’s Website: <http://www.archive.org/details/reportoninsanefe00bill>. Much of the book consists of tabulated data that did not readily convert to a form suitable for screen-reading software. However, the text portions often describe what is of most interest in the tables.