Academic journal article The Western Journal of Black Studies

Can Numbers Lie? Race, Data Collection, and Black and Italian Infant and Maternal Health Care in Early Twentieth Century New York City

Academic journal article The Western Journal of Black Studies

Can Numbers Lie? Race, Data Collection, and Black and Italian Infant and Maternal Health Care in Early Twentieth Century New York City

Article excerpt


In 1899, a young sociologist named W.E.B. DuBois published The Philadelphia Negro, the first sociological study produced in the United States. In it he warns his audience that the best available methods of sociological research are at present so liable to inaccuracies that the careful student discloses the results of individual research with diffidence; he knows that they are liable to error from the seemingly ineradicable faults of the statistical method, to even greater error from the methods of general observation, and, above all, he must ever tremble lest some personal bias, some moral conviction or some unconscious trend of thought due to previous training, has to a degree distorted the picture in his view. Convictions on all great matters of human interest one must have to a greater or less degree, and they will enter to some extent into the most cold-blooded scientific research as a disturbing factor (DuBois 1899, 2-3).

A little more than a century later, sociologists Tukufu Zuberi and Eduardo Bonilla-Silva have taken up the DuBoisian banner by arguing that statistical data can be influenced by the racialized biases of researchers (Zuberi & Bonilla-Silva, 2008). Even when trying to do the right thing, researchers and people who work in public health can generate biased data that directly impacts the lives of persons who often lack the power to persuasively challenge published surveys, position papers, and policy documents. As a result, researchers may find what they're looking for, clothed as "conclusive" data that acts to reinforce the existing societal prejudices that they may be striving to undo. The outcomes are potentially chilling for the fields of public health and medicine, which rely heavily on data collection and modeling for programs, funding, and public policies (Bonilla-Silva & Zuberi 2008; Scott 1997).

This paper builds upon DuBois' admonition and Zubari's and Bonilla-Silva's warnings by examining two public health pamphlets published in 1924 from infant and maternal health experiments, or "demonstrations," that were conducted between 1917-1924 among some of New York City's poorest inhabitants--African American, British West Indian, and Southern Italian women and their children. During research for a book on health work in these areas, I found that, when checked against internal data produced by the same organization that orchestrated the studies and published the results, researchers may have allowed social biases and prejudices to color how one study was conducted, and may have even "sweetened" the published results on another to make their work appear more efficacious. That the areas targeted in a New York City health study conducted in 1915--the first study of its kind through all five boroughs--had been divided in 1893 into racialized "Sanitary Districts" so that critical data on health and disease could be collected by race, was considered normative and necessary at the time (Chaddock, 1916). That researchers and health workers of the New York Association for Improving the Condition of the Poor (AICP), Vanderbilt Clinic and other institutions, empowered by the New York City Department of Health, would create racially-driven health demonstrations and programs for minority women and children, was considered pathbreaking and "the right thing to do" at the time. However, what the data revealed to reading audiences and how it had been created via internal documents, reveal different things.


The decades between 1890-1930 were critical sites for the emergence of many institutions and structures that are today commonplace in the modern worlds of public health and medicine. Statistical studies were new; statistics itself was growing into a separate discipline. The first epidemiological studies themselves were only a few generations old. Bacteriology and publically and privately funded laboratories capable of conducting consistently valid tests were relatively new. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.