Skip to main content
NCBI home page
American Journal of Public Health logoLink to American Journal of Public Health
. 2021 Nov;111(11):1960–1968. doi: 10.2105/AJPH.2021.306454

The Social Survey, the Metropolitan Life Insurance Company, and the Beginnings of the US Public Health Service’s Sickness Surveys

Mónica García 1,✉
PMCID: PMC8630487  PMID: 34709856

Abstract

The earliest sickness survey of the US Public Health Service, which started in 1915, was the Service’s first socioeconomic study of an industrial community. It was also the first to define illness as a person’s inability to work. The survey incorporated the Metropolitan Life Insurance Company’s definition of illness, which, instead of sickness rates, focused on duration of illness as a proxy of time lost from work. This kind of survey took place in the broader context of the reform movements of the Progressive Era and the social surveys conducted in the United States, which led to the creation of the Federal Commission on Industrial Relations, where the Service’s sickness survey originated. The Service’s focus on the socioeconomic classification of families and definition of illness as the inability to work enabled it to show the strong link between poverty and illness among industrial workers. The leader of the survey, Edgar Sydenstricker, and the Metropolitan Life Insurance Company came up with new ways to measure the health of the population, which also influenced the Service’s studies of the effects of the Great Depression on public health and the National Health Survey of 1935–1936. (Am J Public Health. 2021; 111(11):1960–1968. https://doi.org/10.2105/AJPH.2021.306454)

This is a study of the pioneering socioeconomic sickness survey in the United States, undertaken by the US Public Health Service (hereafter referred to as “Service”) in the 1910s. The Service was created in 1912 to assist the public health work of the states. Between 1912 and World War II, the Service carried out socioeconomic investigations, which had an enormous influence on the scientific and public understanding of inequalities in health.1 These investigations also provided the basis for innovative epidemiological designs.2 The Service’s early socioeconomic studies were also an adaptation of the social surveys that expanded after Charles Booth’s work on poverty, Life and Labor of the People in London (1899 and 1903), and Benjamin Seebohm Rowntree’s study of poverty in York, England, A Study of Town Life (1901).3 The social survey in England studied the urban working classes to formulate programs to ameliorate the harsh conditions of their lives.4 The term refers to a number of investigations that, contrary to previous ones, entailed the first-hand collection of data about individuals, families, and households on a local rather than a national level and developed in tandem with the evolution of public policy and social reform.5 Among the most notable were those undertaken by the settlement house movement in England and the United States (i.e., the Hull House Maps and Papers about Chicago, Illinois, 1890), the work of W. E. B. Du Bois (1889), and the surveys sponsored by the Russell Sage Foundation (the Pittsburgh Survey, 1907–1908).6 However, historians have largely ignored the social surveys devoted to public health. The bibliography of social surveys by Allan Eaton and Shelby M. Harrison of the Russell Sage Foundation (1930) shows that public health was indeed at the core of the social survey.7 By focusing on the work of the first economist-statistician of the Service, Edgar Sydenstricker,8 I will show how the Service’s investigations were a continuance of the Victorian and Progressive tradition.

One aim of the social surveys in the United States was to be objective and scientific, which, in the period in question, was synonymous with quantitative methods. Indeed, the Progressive reformers increasingly relied on statistics, economics, and the emerging social sciences: the Service’s early surveys adopted the quantitative analyses typical of the social survey.9 Of course, neither quantification nor the effort to measure the relation between social conditions and health was a novelty at that time. In the 19th century, public health advanced in close connection with the development of statistics and the rationalization of government administration. Starting in 1830, a French doctor, Louis René Villermé, pioneered quantitative investigations that established links between poverty and the poor health of the working classes in Paris, France,10 while in England, the General Register Office, created in 1837, developed an institutional structure for the study of the relationship between mortality, causes of death, and occupation that became the basis of the British public health movement.11 In the United States, during roughly the same period, the decentralization of its government and the lack of a comprehensive compilation of vital statistics meant that there were no systematic studies of mortality patterns in the whole population before the 1930s. However, as Krieger and Fee have pointed out, investigations of the links between socioeconomic conditions and health by federal agencies and nongovernmental bodies were crucial in revealing the socioeconomic factors responsible for inequalities in health in the first half of the 20th century.12

One novel feature of the Service’s surveys, led by Sydenstricker, which quantified the relation between socioeconomic conditions and health, was the employment of the methods pioneered by the Metropolitan Life Insurance Company. As some historians have noted, the insurance companies were responsible for some of the most thorough mortality studies in that period.13 In Britain, the statistician William Farr of the General Register Office used the life tables technique of actuaries to calculate life expectancy, age-specific mortality rates, and standardized mortality rates to determine which districts of the country were healthy or not and improve the conditions of public health, which became the basis of the 20th century British classification of health conditions in terms of occupation.14 To throw further light on the links between US life insurance companies and public health surveys, this article shows how the Service’s investigations implemented the measurements the Metropolitan Life Insurance Company used to investigate the health of its policyholders. The company’s approach not only influenced the Service’s earliest survey but also continued to be used up to the 1930s.

SOCIAL SURVEY AND THE SERVICE

Soon after its creation, the Service came up with a research agenda that would eventually make poverty the main focus of its analyses of the effect of socioeconomic differences on health.15 Sydenstricker was a graduate of Washington and Lee University (1902) and Fellow of Political Economy at the University of Chicago (1907–1908).16 His previous work in labor economics for the Immigration Commission (1907–1909)17 but, most importantly, for the Commission on Industrial Relations (1912–1915), connects Sydenstricker’s work in the Service with the social survey movement initiated in Britain.

The term social survey was derived from Booth’s work. Alarmed by the situation of the poor, certain sectors of the middle and upper classes drafted policies for social reform based on a systematic accumulation of data.18 Quantification was a significant aspect of the approach of Booth and his followers. Rowntree’s work on poverty in York (1901), for example, sought “the true measure of the poverty in the city . . . how much of it was due to insufficiency of income and how much to improvidence.”19 Booth specifically set out to show “the numerical relation which poverty, misery, and depravity bear to regular earnings and comparative comfort” and measure the number of people above and below the poverty line.20 Instead of relying on official statistics, Booth innovatively compiled first-hand data about the household circumstances of poor families gathered by school board visitors. His findings not only had a direct impact on public policies and led to movements for social reform in England, but they also inspired pioneering social researchers in the United States, including Jane Addams and W. E. B. Du Bois.21

The social survey flourished in the United States in the early decades of the 20th century, thanks to the work of the philanthropist Margaret Olivia Sage and the magazine Charities and the Commons (renamed Survey in 1909), a leading national outlet for the reform movement.22 The magazine’s editors, Edward T. Devine and Paul U. Kellogg, with financial support from the Russell Sage Foundation, undertook the Pittsburgh Survey in 1907–1908, which studied how industrial capitalism (the steel industry) shaped urban development.23 After the bombing of the Los Angeles Times building in 1910 and the subsequent conviction and imprisonment of those held responsible for it— John J. McNamara (the secretary of the International Association of Bridge and Structural Iron Workers) and his brother, James McNamara—these editors called on President Taft to create a commission for the investigation of social unrest. The bombing was the culmination of two decades of conflict between capital and labor caused by demands for better working conditions and the right to form trade unions. Along with leaders of the social survey movement, like Addams, the editors argued that the growing power of capitalism in American life was creating unrest in the working class, and because this problem was beyond the capacity of the country’s legal system, a thorough investigation of trade unions and the cost of strikes would be needed to deal with it.24

Thus, pressured by the Progressive reformers and leaders of the US social survey movement, the US Congress created the Commission on Industrial Relations in 1912 to ascertain the causes of industrial strife and ensure decent working and living conditions.25 William Jett Lauck,26 who had studied industrial communities for the Immigration Commission of 1907–1909,27 was appointed director of research. Sydenstricker assisted the sanitary adviser of the Commission on Industrial Relations, the Surgeon General of the Service, Benjamin S. Warren, in the investigation of health insurance for wage earners and the living conditions of industrial workers and communities.28

A staff of more than 50 persons was assembled for the Commission, most of them young economists.29 Chiefly concerned with the strikes and violent protests that took place between 1910 and 1915, they interviewed more than 700 persons all over the country, from businessmen to miners, farmers, garment workers, lumberjacks, silk weavers, and mechanics.30 The Commission served as a training ground for a new generation of labor economists and social investigators who came into prominence during the 1920s and the New Deal.31

Sydenstricker also surveyed community welfare for the Commission and studied the public hearings on the building trades of New York City, collective bargaining in the anthracite coal industry in Pennsylvania, and the conditions of industrial workers.32 In his study of Pennsylvania coal miners, he argued that the causes of industrial unrest were the irregularity of employment, the fall of the real income of the miners, and “the manner of living, standard of home, the real value of the worker’s wages as expressed in commodities and enjoyment”; the wage earner lacked incentives to become a “stable citizen.”33 The unrestrained growth of industrial centers had brought about poverty, physical degeneration, and moral deterioration, he argued.34 Municipal governments, not social workers—with their “taint of charity”—should be responsible for community welfare, in his view.35 His findings supported the Progressive reformers, who wanted the authorities to guarantee the rights of the poor to employment and insurance.36

The Commission believed that the responsibility for the worsening conditions of workers and industrial violence lay with the big corporations: their enormous profits were not fairly shared with their workers or consumers. Instead of tracking real earnings over time, Sydenstrickeŕs report, Conditions of Labor in the Principal Industries, analyzed the extent to which a worker’s income would support a decent standard of life: the average wage-earner’s family spent from three fourths to four fifths of its income on subsistence, leaving little or nothing for unemployment or old age.37 According to Sydenstricker’s report, “Fully one-half of the wage-earners’ families in the United States do not have any income sufficient for adequate subsistence and health.”38

While the Commission had not been formed to investigate the health of workers, in 1915, Warren asked Lauck to help the Service to determine the causes of sickness and mortality among wage earners in a typical industrial community. It rested on Warren’s idea that such problems had complementary “hygienic and economic” causes and, thus, the “economic and hygienic facts should be obtained at the same time and for the same conditions.”39

A one-year study of “a fair type of the American industrial town” of 20 000 to 30 000 inhabitants, with 4000 workers in a single predominant industry and a complete range of wages, would be ideal, argued Warren: the community should be as average as possible,40 and the study should focus on a cross-section of national conditions (although African Americans were excluded).41 The main value of that study, never made before, would be to show “the real” relation of economic and hygienic conditions in a typical industrial community.42

Lauck fully agreed with the “determination in a scientific manner of the relation of sickness and mortality to poverty and all that the term poverty means.” In his opinion, although Rowntree’s study of poverty in York was groundbreaking, it was out of date and only applied to England; because “there is nothing in America of its kind,” it would be “the first survey in the field.”43 Warren claimed that Sydenstricker’s study for the Commission on the cost of living would have a direct bearing on the question of health, a study that “has never been made before;” because the whole subject “so closely involves economic facts and questions,” Sydenstricker was hired by the Service.44

THE SERVICE'S EARLIEST SICKNESS SURVEY

With Sydenstricker on board, Warren carried out his investigation of the causes of sickness and mortality among the wage earners of a typical industrial community. This plan drew on the Service’s studies of pellagra in South Carolina, starting in 1914, under physician Joseph Goldberger. Goldberger sought to determine whether there was a causal relationship between diet and the disease with the use of clinical and experimental methods.45 Warren and Sydenstricker planned to complement Goldberger’s investigations with a study of the economic and sanitary factors responsible for pellagra, not just the dietary ones. For the 1915–1916 study, they chose seven cotton mill villages in northwestern South Carolina, with 500 to 800 inhabitants each, where pellagra was prevalent. A house-to-house canvass was supervised by physician George A. Wheeler, and the assembling of data on the families’ diet and income by Sydenstricker. This field work (which continued until 1919 at least) also investigated the links between income and illness in general, not just with pellagra.

It was the first study of the close correlation of “hygienic” and “economic” conditions in an industrial community46 and drew on the methods of Booth and Rowntree: the first-hand collection, by a trained investigator, of data about individuals, families, and households on a local level, and an attempt to quantify the phenomena and pressure public policy to make social reforms.47 Lauck believed that the Service’s sickness survey would provide a “solid foundation for more intelligent and constructive remedial measures,”48 while Warren and Sydenstricker urged the federal government to implement a system of sickness insurance, arguing that a study of the causes of sickness among wage earners would throw light on the responsibilities of employers, employees, and the public for these conditions.49

Instead of the “indefinite terms” like “poor,” “fair,” and “well-to-do” commonly used to classify the population’s living conditions, Sydenstricker and his associates used quantitative methods and, “for the purpose of accuracy and convenience,” an index of living conditions was expressed in a numerical form.50 They took into account the income and food requirements of the family when classifying the sample population, about 4160 people in about 750 households.51 The resulting groups—half-month income of less than $6.00, $6.00 to $7.99, $8.00 to $9.99, $10.00 to $13.99, and $14.00 or more—showed a crude case rate of pellagra per 1000 of 42.7, 26.0, 12.8, 4.1, and 3.4 for the year 1916, respectively. According to the authors, “It is the first reported study in which the degree of the long-recognized association between poverty and pellagra incidence is measured in a definite, purely objective manner.”52

Similarly, for the analysis of the possible association between family income and sickness in general, the same investigators used the term “disabling illness” as a proxy for sickness to make their results as comparable as possible with the contemporary sickness surveys by insurance companies.53 “Sick” persons were those who were “unable to work” on account of sickness or accident, including those “up and about but unable to work” and those confined to bed.54 The community sickness surveys in question had been made by the Metropolitan Life Insurance Company in 1915. The definition of sickness as disability and the instructions to the field agents of the Service’s survey were exactly those of the insurance company.55 The unemployment surveys the insurance company made in conjunction with the US Bureau of Labor Statistics paved the way for the company’s sickness surveys.56

METROPOLITAN LIFE INSURANCE COMPANY

Unemployment had been one of the major concerns of the investigations of social unrest, not only of the Commission on Industrial Relations but also the Bureau of Labor Statistics and local organizations. The Bureau regarded unemployment as the greatest evil of the capitalist system. The Commission presented a novel official view of unemployment as a collective tragedy that society was responsible for, instead of voluntary organizations or the private sector.57 Meanwhile, the Bureau, the government of New York City, and the Metropolitan Life Insurance Company set out to measure unemployment, first in New York City, then nationwide.58

In January 1915, the agents of the Metropolitan, in conjunction with the Mayoralty of New York, visited the families of the policyholders, 155 960 in all, and concluded that 18% of all potential wage earners were unemployed. Having found the same percentages based on the 1910 census, the Bureau enlisted the insurance company to make studies in other cities. All persons who had any employment, regular or irregular, full or part time, were recorded as employed.59 The Bureau, in cooperation with the insurance company, surveyed 16 cities in the East and Midwest and 12 cities in the Rocky Mountains and the Pacific Coast to find out whether unemployment was because of “no work to be found,” “sickness or disability,” “strikes or lockouts,” or “other causes.”60

Having found that 11% of unemployment was associated with sickness or workplace accidents, the Metropolitan set out to measure the “amount of illness prevailing in American communities.”61 It chose the city of Rochester, New York, for its initial “sickness census” in September 1915. In all, 7638 families, made up of 34 490 persons, were surveyed, almost 10 times the number of families canvassed in the Service’s earliest sickness survey in South Carolina. The census divided sickness into cases in which the illness did or did not result in incapacity (to work). The instructions to the agents specify, for those who were found to be sick, that—

The sick should include: (a) Those persons who are up and about, but are unable to work because of sickness or accident; (b) Those who are confined to bed at home because of disease or accident; (c) Those who are receiving treatment in hospitals or similar institutions.

The question “how long sick to date” should be answered in days, weeks or months62

—the same method of the Service’s earliest community survey.

Disease understood as an inability to work surpassed disease understood as a specific medical diagnosis. Using the former, the company calculated the effects of sickness in the general population of Rochester, in terms of days lost from work and, thus, wages lost in a year. Illness in a community—or the prevalence of sickness—could thus be translated into economic losses—and perhaps into calculating premiums.63

The company noted that “a sick benefit society experiencing a light sickness rate may come to grief if the sicknesses involve protracted payments of benefit out of all relation to the premiums charge.” So, classifying sickness rates by sex, age, and other measures was perhaps less important than

to know that, of those cases of sickness, 56% will probably last more than one month, 39% more than 3 months, 25% more than one year and 14% more than 3 years . . . persons interested in the social consequences of sickness are more likely to inquire into the durations of sicknesses than, primarily, into the sickness rate.64

This definition of sickness dates back to the company’s free Visiting Nurse Service created in 1909. Lee K. Frankel, then head of the company’s welfare and health program, believed that the nurse service benefitted both the policyholder, usually a worker, and the company: the policyholders who were taught the importance of personal hygiene would be less likely to suffer from illness and premature death, and the company would likewise profit from a lower mortality rate.65 But the company’s need to show that spending a policyholder’s money on health work would reduce the cost of insurance, with findings backed by data,66 led it to compile information about the policyholders, in the form of a “new case and history slip” the nurses had to fill in at the bedside, which included such questions as “how long was the patient ill at the time of the first visit? Up and about? In bed?”67 This information was grouped into the general category of “inability to work” in the sickness surveys done in Rochester in 1915, which in turn were taken up by the Public Health Service’s sickness survey in the cotton mill villages in South Carolina. With this definition and metrics, the Service found in 1915 to 1916 that the rates of disabling illness (inability to work) per 1000 persons in the families of the cotton mill villages with an income of less than $6.00, $6.00 to $7.99, $8.00 to $9.99, and $10.00 and more were 70.1, 48.2, 34.4, and 18.5, respectively.68

THE LATER SICKNESS SURVEYS

The socioeconomic classification of the families used by Sydenstricker in the South Carolina sickness survey, and the definition of illness proposed by Metropolitan Life Insurance Company, were at the core of several sickness surveys of the following decades. In the Service’s investigation of the mortality rates from influenza during the pandemic of 1918 and the sickness survey of Hagerstown, Maryland (1921–1924), of the incidence of several diseases among 1600 families (7200 persons), information on the economic status of the respondents was recorded but it was left to the enumerators to classify the families with the use of “indefinite”—not quantitative—categories: “well-to-do, comfortable, moderate, poor, very poor.”69 The 1928 and 1931 surveys of 8758 families in 17 states by the Committee on the Cost of Medical Care, with the participation of Sydenstricker and the insurance company, used the company’s definition of illness: “any disorder which wholly or partially disables an individual for one or more days. . . .”70; and the number of days in bed, days lost from school or work, and total duration of each illness were also recorded.71

In the studies done during the Depression by the Milbank Memorial Fund and the Service with the Metropolitan Life Insurance Company under the auspices of the Health Organization of the League of Nations, quantitative data of 12 000 families in 10 localities, including eight large cities, were compiled. The households were classified according to their economic status as comfortable, moderate, or poor to determine and quantify whether a fall in income had affected their health.72 A “disabling illness” was defined as a situation in which “the person was unable to work, attend school or undertake other activities for 1 or more days during the 3 months period of the study.”73 This methodology was based on the assumption that mortality statistics were an insufficient way to measure a nation’s health. The conservative members of the League of Nations Health Organization, who eventually won the upper hand, had claimed that the decline in mortality rates during the 1930s meant that the economic crisis had not affected public health, while Sydenstricker, now working for the Milbank Memorial Fund, and George St John Perrott, from the Service, showed the opposite: that morbidity (disabling illness) was directly linked with socioeconomic status (income).74

Historian George Weisz has shown that defining disabling illness as inability to work and perform other activities was crucial for the focus on chronic disease of the biggest morbidity survey up to then: the National Health Survey (1935‒1936), whose aim was to support health care reform.75 Sydenstricker was the driving force of this survey as well. Weisz argued that its focus revealed the pervasiveness of chronic disease and how the poor, who had less access to health care, suffered more from diseases than other classes.76

CONCLUSIONS

The earliest socioeconomic sickness survey in an industrial community undertaken by the US Public Health Service in 1915, a result of the US social survey movement, marked the beginning of a methodology of research that stretched to the National Health Survey of the 1930s. Other agencies, like the US Children’s Bureau and the National Tuberculosis Association, also used information on a family’s economic status or occupation in their mortality investigations. With the decline of the Progressive movement, institutional and methodological changes, and the outbreak of World War II, the socioeconomic measurements used for vital statistics were virtually abandoned.77 However, the 1935–1936 National Health Survey’s definition of illness in terms of days of disability played a significant role in the United States’ epidemiological and political approach to chronic diseases, while the National Health Survey became a major source of information for the arguments that eventually led to the establishment of Medicaid.78 This definition originated with the Metropolitan Life Insurance Company’s sickness surveys and was adopted by the surveys that the Public Health Service began in the 1910s. Both organizations participated in the Committee on the Cost of Medical Care and in the studies made during the Depression. The extent to which the Service’s concerns about the health of the wage earners coincided with the interests the company is a subject that requires further research.

ACKNOWLEDGMENTS

Research for this article was made possible thanks to the financial support of the Universidad del Rosario (Colombia) for research at the US National Archives at College Park, MD, and for a visit to the Department of Social and Behavioral Sciences at the T. H. Chan School of Public Health‒Harvard University, during the fall of 2018.

 I would like to thank Andrés Jiménez Angel, James Weiskopf, and the anonymous referees whose comments greatly improved the article. I am also grateful to the archival staff of the National Archives and the Harvard Libraries for invaluable help in tracking down sources.

CONFLICTS OF INTEREST

The author reports no conflict of interest.

ENDNOTES

  • 1.Krieger Nancy, Fee Elizabeth.Measuring Social Inequalities in Health in the United States: A Historical Review, 1900‒1950 International Journal of Health Services 2631996391–418. 10.2190/B3AH-Q5KE-VBGF-NC74. [DOI] [PubMed] [Google Scholar]
  • 2.Mooney Stephen J., Knox Justin, Morabia Alfredo. American Journal of Epidemiology 180, no. 3 (2014): 235‒244, 10.1093/aje/kwu134; John M. Eyler, “Health Statistics in Historical Perspective,” in Health Statistics: Shaping Policy and Practice to Improve the Population’s Health, eds. Daniel J. Friedman, Edward L. Hunter, and R. Gibson Parrish (New York, NY: Oxford University Press, 2005), 25‒52; Paul Weindling, “Interwar Morbidity Surveys: Communities as Health Experiments,” in Facing Illness in Troubled Times, eds. Iris Borowy and Wolf D. Gruner (Frankfurt am Main, Germany: Peter Lang, 2005), 75‒84. [DOI]
  • 3.Bulmer Martin, Bales Kevin, Sklar Kathryn Kish. The Social Survey in Historical Perspective. In: Bulmer Martin, Bales Kevin, Sklar Kathryn Kish., editors. The Social Survey in Historical Perspective, 1880‒1940. Cambridge, England: Cambridge University Press; 1991. pp. 1–48. [Google Scholar]
  • 4.Bulmer, et al. , “The Social Survey in Historical Perspective,” 48; Eileen Janes Yeo, “Social Surveys in the Eighteenth and Nineteenth Centuries,” in The Cambridge History of Science, Volume 7, The Modern Social Sciences, eds. Theodore Porter and Dorothy Ross (Cambridge, England: Cambridge University Press, 2003), 83‒99.
  • 5.Bulmer, et al.
  • 6.On social survey, see Maurine W. Greenwald and Margo Anderson, Pittsburgh Surveyed. Social Science and Social Reform in the Early Twentieth Century (Pittsburgh, PA: University of Pittsburgh Press, 1996); for the Settlement House movement in the United States, see, for example, Harry P. Kraus, The Settlement House Movement in New York City, 1886‒1914 (New York, NY: Arno Press, 1980).
  • 7.Eaton Allen, Harrison Shelby M. A Bibliography of Social Surveys. New York, NY: Russell Sage Foundation; 1930. [Google Scholar]
  • 8.The primary sources upon which this research was based are documents available at the National Archives at College Park, MD, and Harvard Libraries. At the National Archives, I checked the papers of the Commission on Industrial Relations where Sydenstricker worked before going to the US Public Health Service. They are located in the Microfilm Publications, T4, A2. These documents have been used by economic historians such as Mary O. Furner but never in relation to the Service’s surveys. I also checked the papers and correspondence of the pellagra studies directed by the Service, where surgeon Benjamin S. Warren (advisor of the Service to the Commission on Industrial Relations) and Edgar Sydenstricker started their socioeconomic surveys. They are available at RG Central File 1897‒1923, Pellagra, NC-34-10, File 1648, Boxes 148‒153. I also checked the correspondence of the Service in RG 90 Insurance-Health, which refers to the initiative to make community sickness surveys and how it was related to the social survey. Finally, I also checked the National Institute Health Survey Forms, Hagerstown, in RG 90 Entry: A1 21, Boxes 1-20. As for the social survey and the Metropolitan Life Insurance Company, I relied on the historiography about the company as well as on published primary sources located at the Countway Library of Medicine and the Widener Library, both at Harvard University.
  • 9.Bender Thomas, A Nation Among Nations. America’s Place in World History (New York, NY: Hill and Wang, 2006), particularly chapter 5, “The Industrial World and the Transformation of Liberalism,” 246‒295; Richard L. McCormick, “Public Life in Industrial America, 1877‒1917,” in Eric Foner, The New American History (Philadelphia, PA: Temple University Press, 1997), 107‒132; Greenwald and Anderson, Pittsburgh Surveyed; Graham Adams Jr, Age of Industrial Violence, 1910‒1915 (New York, NY, and London, England: Columbia University Press, 1966); Mary O. Furner, Advocacy and Objectivity: A Crisis in the Professionalization of American Social Science, 1865‒1905 (Lexington, KY: University Press of Kentucky, 1975).
  • 10.Julia Chantal, Valleron Alain-Jacques.Louis-René Villermé (1782‒1863), a Pioneer in Social Epidemiology: Re-analysis of His Data on Comparative Mortality in Paris in the Early 19th Century Journal of Epidemiology and Community Health 652011666–670. 10.1136/jech.2009.087957. [DOI] [PubMed] [Google Scholar]
  • 11.Szreter Simon, “The GRO and the Public Health Movement in Britain, 1837‒1914,” Social History of Medicine 4, no. 3 (1991): 435‒463, 10.1093/shm/4.3.435; and Simon Szreter, “The Official Representation of Social Classes in Britain, the United States, and France: The Professional Model and ‘Les Cadres,’” Comparative Studies in Society and History 35, no. 2 (1993): 285‒317. [DOI]
  • 12.Krieger and Fee, “Measuring Social Inequalities in Health,” 394–395..
  • 13.Eyler, “Health Statistics in Historical Perspective,” 25‒52, 39.
  • 14.Szreter, “The GRO and the Public Health Movement,” 435‒463; and Szreter, “The Official Representation of Social Classes,” 285‒317.
  • 15.Krieger and Fee, “Measuring Social Inequalities in Health,” 394–399..
  • 16.Sydenstricker is listed as one of the graduate students of the University of Chicago in James Lawrence Laughlin, Twenty-Five Years of the Department of Political Economy (University of Chicago, Priv. Print., 1916), 16. See also Wilford I. King, “Edgar Sydenstricker,” Journal of the American Statistical Association 31, no. 94, 1936: 411–415; Washington and Lee University, The Calix 1902. Yearbook, Vol VII (Louisville, KY: The Prentice Press, 1902), 25. What we know about Edgar Sydenstricker derives from his work for several institutions. For his work for the Commission on Industrial Relations, see Mary O. Furner, “Knowing Capitalism: Public Investigation and the Labor Question in the Long Progressive Era,” in The State and Economic Knowledge. The American and British Experiences, eds. Mary O. Furner and Barry Supple (Cambridge, England: Cambridge University Press, 1990), 275. About his work for the US Public Health Service, see Krieger and Fee, “Measuring Social Inequalities in Health,” and Harry Marks, “Epidemiologists Explain Pellagra: Gender, Race and Political Economy in the Work of Edgar Sydenstricker,” Journal of the History of Medicine 58, no. 1 (2003): 34‒55, https://doi.org/10.1093/jhmas/58.1.34. We also know that he assisted the League of Nations Health Organization as a statistical adviser: see Weindling, “Interwar Morbidity Surveys,” 77. He also helped in the design of the National Health Survey of 1935‒1936: see George Weisz, “Epidemiology and Health Care Reform. The National Health Survey of 1935‒1936,” American Journal of Public Health 101, no. 3 (2011): 438‒447, https://doi.org/10.2105/AJPH.2010.196519. Finally, we also know that while working for the Milbank Memorial Fund, he advocated for national health insurance before the Committee on the Cost of Medical Care and the Committee on Economic Security of the Roosevelt administration: see Jaap Kooijman, . . . And the Pursuit of National Health. The Incremental Strategy Toward National Health Insurance in the United States of America (Amsterdam, Netherlands: Rodopi, 1999).
  • 17. Sydenstricker’s work for the Immigration Commission was most likely his first experience of fieldwork on the condition of industrial workers and the use of quantitative methods and descriptive accounts in industrial surveys. The purpose of the Immigration Commission was to study the origins of the immigration of the 1880s, motivated by the higher wages paid for industrial labor and its effects upon the institutions and industries of the country—see Reports of the Immigration Commission. Abstracts of Reports of the Immigration Commission: With Conclusions and Recommendations and Views of the Minority, V1 (Washington, DC: Government Printing Office, 1911), 13. The Commission reflected the spirit of Progressivism, a movement to improve the country’s socioeconomic conditions, relying on what their adherents considered to be scientific objectivity and a commitment to government intervention. The Progressives believed that investigations carried out by properly trained experts, not by well-meaning amateurs or charities, would equip policymakers with the means to solve social problems—see Robert F. Zeidel, Immigrants, Progressives, and Exclusion Politics. The Dillingham Commission, 1900‒1927 (DeKalb, IL: Northern Illinois University Press, 2004), 3. Two of the nine-member Commission were professors of political economy, and the support staff included the political economist William Jett Lauck, who was trained by James Laurence Laughlin and Thorstein Veblen at the Department of Political Economy of the University of Chicago—see Laughlin, Twenty-Five Years of the Department of Political Economy, 9. One of Lauck’s assistants was Edgar Sydenstricker, then a graduate student of political economy at the University of Chicago—see Laughlin, Twenty-Five Years of the Department of Political Economy, 16; see also King, “Edgar Sydenstricker.” The Commission obtained original information on about 3.2 million people. Their fieldwork was done in 24 states with 17 141 households and studied several industries, from sugar refining to coal mining. The investigations combined data “which were susceptible of tabulation and statistical presentation” about the individual employee, his household, and industrial plant, along with descriptive and historical information about 200 industrial plants and communities: see Reports of the Immigration Commission, 293‒295.
  • 18. These sectors were the “professional classes,” that is, women of a privileged social position, professionals in the field of medicine, members of voluntary welfare associations, some journalists, and, at the end of the century, a few academics responsible for “social surveys.” See Bulmer et al., “The Social Survey in Historical Perspective,” 2.
  • 19.Rowntree Benjamin Seebohm.Poverty. A Study of Town Life London, England: Macmillan and Co Limited; 1901vi. [Google Scholar]
  • 20.Booth Charles.Life and Labour of the People in London. Volume I. East, Central and South London London, England: Macmillan and Co; 18926. [Google Scholar]
  • 21.Bulmer, et al.
  • 22.Converse Jean M., Survey Research in the United States. Roots and Emergence, 1890‒1960 (New Brunswick, NJ: Transactions Publishers, 2009), 22‒23; Martin Bulmer, “The Decline of the Social Survey Movement and the Rise of American Empirical Sociology,” in Bulmer et al., The Social Survey in Historical Perspective, 991–315, 293.
  • 23.Cohen Steven R. The Pittsburgh Survey and the Social Survey Movement: A Sociological Road Not Taken. In: Bulmer, et al., editors. The Social Survey in Historical Perspective. pp. 245–268. [Google Scholar]
  • 24.Adams, Age of Industrial Violence, 25‒30.
  • 25.First Annual Report of the Commission on Industrial Relations Chicago, IL: Barnard & Mills Print; 19146. [Google Scholar]
  • 26. Lauck graduated from Washington and Lee University and obtained a degree in political economy at the University of Chicago. See Laughlin, Twenty-Five Years of the Department of Political Economy, 9; Washington and Lee University, “University, Washington, D.C.,” The Ring-Tum Phi 11, no. 17 (1908): 2; Washington and Lee University, “Hart, Schaffner & Marx Prizes,” The Ring-Tum Phi 12, no. 11 (1909): 4.
  • 27. Washington and Lee University, “Dr G. Campbell Tells of Industrial Commission. Vitally Important Work Being Done by W. & L. Men,” The Ring-Tum Phi 17, no. 18 (1914): 3.
  • 28.Warren Benjamin S., “Outline of Plan for the Study of Sickness Insurance,” April 23, 1915, in National Archives, RG 90 Insurance-Health, 6. Elizabeth Fee’s and Theodore M. Brown’s assertion that the available biographical information about Benjamin Warren is very meager still holds true: see Elizabeth Fee and Theodore Brown, “Edgar Sydenstricker and Benjamin Warren,” American Journal of Public Health 89, no. 11 (1999): 1643–1644, 1643. Warren’s participation in some of the Service’s projects, such as the pellagra studies, is mentioned in Ralph Chester Williams, The United States Public Health Service, 1789‒1950 (Washington, DC: US Public Health Service, 1951), 152, 423. To my knowledge, this article is the first to discuss his work as the sanitary adviser of the Commission on Industrial Relations and his role in implementing the sickness survey among wage earners.
  • 29.First Annual Report of the Commission, 8.
  • 30. Adams, Age of Industrial Violence, 206; Reports of the Immigration Commission, 2.
  • 31.Furner Mary O. Knowing Capitalism: Public Investigation and the Labor Question in the Long Progressive Era. In: Furner Mary O., Supple Barry., editors. The State and Economic Knowledge. The American and British Experiences. Cambridge, England: Cambridge University Press; 1990. pp. 241–286. [Google Scholar]
  • 32.Sydenstricker Edgar, “Topical Analysis of Testimony at Public Hearings on the Building Trades of New York City, N.Y. May 25‒29, 1914,” in Microfilm publications, T4, Reel 14, A2 Cab 41/4, 50 pages; Edgar Sydenstricker, “Report on the System of Collective Bargaining in the Anthracite Coal Industry,” in National Archives, Microfilm Publications, T4, A2 Cab 41/4, Reel 8. See also Edgar Sydenstricker, “Condition of Labor in the Principal Industries,” s/f, in National Archives, Microfilm Publications, T4, A2 Cab 41/4, Reel 9. This served as the basis for W. Jett Lauck and Edgar Sydenstricker, Conditions of Labor in American Industries (New York, NY, and London, England: Funk & Wagnalls Company, 1917).
  • 33.Sydenstricker Edgar.
  • 34.Sydenstricker Edgar.
  • 35.Sydenstricker, “Welfare Activities,” 2‒3; H. G. Lee, “Digest of Mr. Sydenstricker’s Preliminary Report on Welfare Activities of Communities,” February 18, 1915, in National Archives, Microfilm Publications, T4, A2 Cab 41/4, Reel 11.
  • 36.Social progressives, under the American Association for Labor Legislation, played an important role in establishing workmen’s compensation and the campaign for health insurance. See Paul Star, The Social Transformation of American Medicine (New York, NY: Basic Books, 1982), 243‒249; Kooijman, . . . And the Pursuit of National Health, 25‒30.
  • 37.Furner, “Knowing Capitalism,” 278.
  • 38.Sydenstricker, “Condition of Labor in the Principal Industries,” 1.
  • 39.Warren Benjamin S.
  • 40.To ensure a fair sampling—or the “representative method” as it was called—the American social surveys based their selection of communities on the idea of a “typical” village or “average” community, as Charles Booth did when he used the School Board Visitors’ data on the poor in London, England, and the Service when it surveyed the cotton-mill villages in South Carolina. On the representative method in US social survey, see Alain Desrosières, “The Part in Relation to the Whole: How to Generalize? The Prehistory of Representative Sampling,” in Bulmer et al., eds., The Social Survey in Historical Perspective, 217‒244, 220.
  • 41.Warren excluded African Americans because he wished to model his survey on an “average” community; he argued that too great a diversity of races among the wage earners should be avoided because of variations in standards of living caused by “racial customs.” He was also concerned about the impact of his conclusions: he wanted to avoid any “appearance of muckraking,” on the one hand, or present better-than-ordinary conditions, on the other. For him, “complicating” factors should be avoided; hence, as far as possible, the study had to be based on typical conditions. See Warren, “Industrial Hygiene,” 2, 8. Nevertheless, historian Harry M. Marks has shown that Sydenstricker’s focus on economic underdevelopment in the Service’s pellagra studies led him to ignore that African Americans were the main victims of the disease. This indifference, he argues, also reflected a broader concept of race at that time. For Northerners, race meant “immigrants.” When settlement workers and the social economists of Sydenstricker’s cohort spoke of “races,” they were generally referring to recent immigrants. See Marks, “Epidemiologists Explain Pellagra,” 53‒55.
  • 42.Warren, “Industrial Hygiene,” 2, 8.
  • 43.Letter by Jett Lauck to Benjamin S. Warren, Chicago IL, February 17, 1915, National Archives RG 90 Insurance, Health, 1‒2.
  • 44.Warren Benjamin S.
  • 45.Mooney, et al.
  • 46.Letter by GA Wheeler, P.A. Surgeon, Medical Officer, in Immediate Charge, to Surgeon General, USPH Service, Field Investigations of Pellagra, Spartanburg, SC, December 10, 1919, National Archives, RG 90 Central File 1897-1923, Pellagra, NC-34 10, File 1648, Box 148 to 153, Box 151; Letter by Joseph Goldberger, Surgeon in Charge of Pellagra Investigations, to Surgeon General, USPH Service, Washington, July 2, 1920, National Archives, RG 90 Central File 1897‒1923, Pellagra, NC-34 10, File 1648, Box 148 to 153, Box 151.
  • 47.Sydenstricker and his collaborators referred to Seebohm Rowntree’s work on poverty in York, England, as a “classic” when discussing the measurement of the economic status of families: Edgar Sydenstricker and Wilford I. King, “A Method of Classifying Families According to Incomes in Studies of Disease Prevalence,” Public Health Reports 35 no. 48 (1929): 2829‒2846, 2833; and Edgar Sydenstricker and Wilford I. King, “The Measurement of the Relative Economic Status of Families,” Quarterly Publications of the American Statistical Association 17, no. 135 (1921): 842‒857, 844. They even used the information gathered by the Service’s earliest sickness survey in South Carolina cotton-mill villages to ascertain the variations in economic status at different stages of family life, as Rowntree had done in his book and as a direct application of Rowntree’s idea: Edgar Sydenstricker, Willford I. King, and Dorothy Wiehl, “The Income Cycle in the Life of the Wage-Earner,” Public Health Reports 39, no. 34 (1924): 2133‒2140.
  • 48.Letter by Jett Lauck to Benjamin S. Warren, Chicago IL, February 17, 1915, National Archives, RG 90 Insurance, Health, 1.
  • 49.Warren Benjamin S., Sydenstricker Edgar.Health Insurance. Its Relation to the Public Health Public Health Bulletin 2nd edWashington, DC: Government Printing Office; 191638. [Google Scholar]
  • 50.Sydenstricker Edgar, King Wilford I. A Method of Classifying Families According to Incomes in Studies of Disease Prevalence. Public Health Reports. 1920;35(48):2829–2846. [Google Scholar]
  • 51.The size, ages, and genders of the families were taken into account by employing a common denominator for individuals, derived from the Atwater scale of food requirements. Thus, Sydenstricker and his colleagues set 1 as the requirement of a male adult, while the requirements of children and females were set at between 0.2 and 0.8. The 747 households were then classified according to a half-month family income per adult male unit with data canvassed between April and June, 1916. Joseph Goldberger, G. A. Wheeler, and Edgar Sydenstricker, “A Study of the Relation of Family Income and Other Economic Factors to Pellagra Incidence in Seven Cotton-Mill Villages of South Carolina in 1916,” Public Health Reports 35, no. 46 (1920): 2673‒2714, 2678‒2684.
  • 52.Goldberger, et al.
  • 53.Sydenstricker Edgar, Wheeler G. A., Goldberger Joseph.Disabling Sickness Among the Population of Seven Cotton Mill Villages of South Carolina in Relation to Family Income Public Health Reports 334719182038–2051.. [Google Scholar]
  • 54.Sydenstricker, et al.
  • 55.John Eyler also mentions this connection in his chapter “Health Statistics in Historical Perspective,” 39‒40.
  • 56.Frankel Lee E., Dublin Louis I. Community Sickness Survey. Rochester, N.Y., September 1915. Public Health Reports. 1916;31(8):423–438. [Google Scholar]
  • 57.Furner, “Knowing Capitalism,” 279. On the changing views on unemployment since the 19th century, see Ingrid Liebeskind Sauthier, “Modern Unemployment: From the Creation of the Concept to the International Labour Office’s First Standards,” in Globalizing Social Rights. The International Labour Organization and Beyond, eds. Sandrine Kott and Joëlle Droux (Basingstoke, England, and New York, NY: Palgrave Macmillan, 2013), 67‒84; Noel Whiteside, “Constructing Unemployment: Britain and France in Historical Perspective,” Social Policy & Administration 48, no. 1 (2014): 67‒85, https://doi.org/ 10.1111/spol.12032.
  • 58.US Bureau of Labor Statistics Unemployment in New York City, N.Y.,” Bulletin of the US Bureau of Labor Statistics 17219156. [Google Scholar]
  • 59.US Bureau of Labor Statistics
  • 60.US Bureau of Labor Statistics Unemployment in the United States Bulletin of the US Bureau of Labor Statistics 195191692. [Google Scholar]
  • 61.Frankel and Dublin. “Community Sickness Survey,” 423. According to John Eyler, between 1915 and 1917, the Metropolitan Life Insurance Company undertook similar surveys in seven communities in total: five large cities and two smaller villages, where study groups were asked to fill out a simple form about illness for each family they visited. The surveys gathered information on 637 000 individuals. See Eyler, “Health Statistics in Historical Perspective,” 39‒40. See also Louis I. Dublin, A Family of Thirty Million: The Story of the Metropolitan Life Insurance Company (New York, NY: Metropolitan Life Insurance Company, 1943), 438.
  • 62.Frankel and Dublin, “Community Sickness Survey,” 423‒425.
  • 63.Ibid., 438.
  • 64.Frankel Lee K., Dublin Louis I. Sickness Survey of Principal Cities in Pennsylvania and West Virginia. Sixth Community Sickness Survey. New York, NY: Metropolitan Life Insurance Company; 1917. [Google Scholar]
  • 65.Dublin Louis I., After Eighty Years: The Impact of Life Insurance on the Public Health Gainesville, FL: University of Florida Press; 196638–41.. [Google Scholar]
  • 66.Call of the Life Extension Committee of the Association of Life Insurance Presidents, in 1912. F. W. Jenkins, “Report of Health Committee of Association of Life Insurance President,” Proceedings of the Annual Meeting of the Association of Life Insurance Presidents 6 (1912): 78‒85, 83, quoted in Dan Bouk, How Our Days Became Numbered, Risk and the Rise of the Statistical Individual (Chicago, IL: The University of Chicago Press, 2015), 130‒131, footnote 61.
  • 67.Bouk, How Our Days Became Numbered, 131‒132. Bouk’s book includes reproductions of the slips on pages 132 and 133.
  • 68.Income per adult male unit: see Sydenstricker et al., “Disabling Sickness,” 2038‒2039; 2043‒2046.
  • 69.Morabia Alfredo, “The US Public Health Service House-to-House Canvass Survey of the Morbidity and Mortality of the 1918 Influenza Pandemic,” American Journal of Public Health 111, no. 3 (2021): 438‒445, 10.2105/AJPH.2020.306025 [DOI] [PMC free article] [PubMed]
  • 70.Falk Isidore S., Klem Margaret C., Sinai Natham.The Incidence of Illness and the Receipt and Costs of Medical Care Among Representative Families. Experiences in Twelve Consecutive Months During 1928‒1931 Chicago, IL: The University of Chicago Press; 19338. [Google Scholar]
  • 71.Here, the notion of “time lost from work,” although recorded, was considered misleading as measurement of disability attributable to sickness because the interruption of the normal routine depended on whether or not the individual was employed or unemployed at the time. Thus, this factor was eliminated by restricting the index to days in bed and to those who were employed at some time during the survey year. See Falk et al., “The Incidence of Illness,” 81.
  • 72.Perrott George St. John, Collins Selwyn D., Sydenstricker Edgar, “Sickness and the Economic Depression. Preliminary Report on Illness in Families of Wage Earners in Birmingham, Detroit, and Pittsburgh,” Public Health Reports 48, no. 41 (1933): 1251‒1264, 1255‒1259; George St. John Perrott and Selwyn Collins, “Relation of Sickness to Income and Income Change in 10 Surveyed Communities,” Public Health Reports, 50 (1935): 595–622, 607.
  • 73.Perrott, et al.
  • 74.For an analysis of this controversy see Weindling, “Interwar Morbidity Surveys,” 75–84; Iris Borowy, “Crisis as Opportunity: International Health Work During the Economic Depression,” Dynamis, 28 (2008): 29–51, https://doi.org/ 10.4321/s0211-95362008000100002; Mónica García, “Mortality Rates or Sociomedical Indicators? The Work of the League of Nations on Standardizing the Effects of the Great Depression on Health,” Health Policy and Planning 29, no. 1 (2014), 1‒11, 10.1093/heapol/czs111. [DOI]
  • 75. Weisz, “Epidemiology and Health Care Reform.”.
  • 76.Ibid., 445.
  • 77.Krieger and Fee, “Measuring Social Inequalities in Health,” 406‒413.
  • 78.Weisz, “Epidemiology and Health Care Reform,” 446.

Articles from American Journal of Public Health are provided here courtesy of American Public Health Association

RESOURCES