The Man Who Predicted Pandemic Inequality a Century Ago
Edgar Sydenstricker dug into the pandemic of 1918 and found income level was a key factor in who lived and who died
Edgar Sydenstricker, the preeminent epidemiologist of his generation, showed how the 1918 flu pandemic was hardest on the poor, as were so many of the health conditions he studied. He had a plan to fix it. Today, we’re finding, and debating, the same things.
As the flu pandemic of 1918 circled the United States in wave after wave or two years, the poor died in droves. It was a predictable outcome. Researchers in the burgeoning field of epidemiology had just documented similar patterns with tuberculosis, and pointed to obvious causes — comorbidities resulting from hard physical labor, overwork, and poverty, and transmission within overcrowded, substandard housing.
But someone had to do the work to document it. Edgar Sydenstricker had an ideal education for the time, and a perfectly Dickensian name for such a figure. The young researcher began as a labor economist, switching to health economics at the dawn of the field in the U.S. He arrived at the U.S. Public Health Service (PHS) after studying income and working conditions during a particularly brutal period, just in time for the greatest health crisis in the country’s history.
What he found resonates today; we’re much better at gathering the data, but not necessarily acting on it. And like many peers a century later, he found himself not only using the tools of epidemiology and economics to figure out what had happened, but applying them forward, and applying European models to perhaps do things better.
Edgar Sydenstricker was born into a studious family. His sister Pearl S. Buck, author of the high-school staple The Good Earth, was awarded the Nobel Prize in literature in 1938. His other sister, Grace Yaukey, wrote more than 30 books, including two about Pearl. Their father was a Presbyterian missionary to China and the author of a book about the idioms of Mandarin.
Sydenstricker would translate the world into numbers. After a brief stint in something like the family business as a journalist in central Virginia, he studied economics at the University of Chicago and Johns Hopkins. They were arguably the best schools in the country at the time to produce a health economist, one the birthplace of modern sociology and the other the birthplace of modern infectious disease research. Soon he was hired as a special investigator for the U.S. Immigration Commission and the U.S. Commission on Industrial Relations, surveying wages and working conditions, particularly in immigrant-heavy industries. He started this research in 1908, two years after Upton Sinclair published The Jungle, about the immigrant-heavy meatpacking industry in the city Sydenstricker became an economist. Wages and conditions for the population Sydenstricker was studying tended to be bad.
In 1915, he became the PHS’s first statistician, and his first job was to do his last job — study conditions in the New York garment industry, four years after it produced one of the worst and most transformative industrial disasters in U.S. history, the Triangle Shirtwaist Factory fire, in which 146 people, mostly young women, were killed. Conditions were, of course, bad, but “data showing this objectively are rare in the literature,” a colleague wrote by way of introducing Sydenstricker’s work. His survey showed that a quarter of the poorest workers in his sample suffered from “poor” nutrition, with 10% anemic, and 20% mortality among their children.
His next job was to figure out what to do about it.
With Benjamin S. Warren, a surgeon at the PHS who had previously studied the relationship between nutrition and pellagra and would go on to lead the PHS, he produced Health Insurance: Its Relation to the Public Health. Its first line: “The growing realization of the fact that the health of the wage-working population depends on large measure upon economic conditions is leading to the conviction that there is need for more comprehensive measures for the relief and prevention of disease.”
Its second paragraph suggested the measure: “Health insurance has been adopted in many European countries as the remedy for similar conditions and has become an efficient measure for the relief of sickness and an important agency in the prevention of disease. Recent discussions and proposals of health insurance measures are beginning to focus public attention on this subject and to suggest that a governmental system of health insurance is the solution of the problem in America.” (Warren and Sydenstricker also wrote favorably of the concept of a “living wage” as a health policy measure.)
Warren and Sydenstricker identified a number of these economic conditions and their intersection with health. Low wages led to excessive hours in physically demanding industries, but even worse was inconsistent, seasonal piecemeal work, where laborers were paid by the piece instead of by the hour or week, often in short-term gigs, pushing them towards an even greater pace. After the job ended, the money ran out soon enough. Wageworkers would grow old, “in the physiological sense of the term, earlier than persons engaged in other pursuits,” growing less able to perform industrial tasks, and slip down the socioeconomic ladder, worsening their interlocking problems.
The end result? “Modern industry has little use for the man over 45 years of age.”
Warren and Sydenstricker also addressed women as wageworkers, finding greater morbidity — more precarious employment meant they started their days earlier and pushed harder — and an exit from the workplace a decade earlier than men. They concluded that factory work had a negative effect on infant health, but “poverty has a much more deleterious influence, and if by employment poverty can be removed or lessened, such employment is lesser by far of the two evils.”
Poverty wages also created unhealthy living conditions, which would become deadly in the pandemic shortly to come. The U.S. Federal Immigration Commission, where Sydenstricker got his start, found that in nearly a third of newer immigrant households all but one room in their house or apartment was used as a bedroom, and in nearly a third a “separate family existence” was blocked by the need to house lodgers, generally four to 10 among the newest waves of Eastern European immigrants. A literature review found considerably higher death rates from tuberculosis in small and overcrowded homes compared to larger spaces. The worst conditions were found in lodging houses, where a dozen people might live in a small house, or even a room. In New York, the risk of lodging-house residents for tuberculosis was 11 times higher than average.
(Another thing that hasn’t changed much: “The great size of New York City and its proportionally great number of ill-fed and poorly housed working people have naturally concentrated attention on the problem there more than in other centers; but it has been found that living conditions are as bad in smaller industrial centers and in some instances appreciably worse… Furthermore, it should be remembered that the great majority of American wage earners live in the middle-sized and smaller places.” So there.)
Two years later a pandemic tore through the populations Warren and Sydenstricker described, and Sydenstricker followed in its path. The 1918 flu pandemic — known as the “Spanish flu,” but likely originating in the Great Plains — would gain a reputation for killing the rich and poor alike, but Sydenstricker would find that it was worst for the poor, working class, and immigrants, as Covid-19 would prove a century later, for nearly identical reasons.
The best evidence we have about the 1918 flu suggests it was quickly spread by conditions that simulated those Sydenstricker had previously found in his research on immigrants: the conditions of war. The first outbreak on record occurred in Haskell County, Kansas in February 1918, infecting a large share of the sparsely populated farm community. That almost certainly seeded the much worse outbreak at Camp Funston, which drew soldiers from Haskell County. As John Barry writes in his book The Great Influenza, a record-breaking cold winter led to “overcrowded and inadequately heated” barracks — the army’s words, not Barry’s — and an insufficient supply of warm clothing led to the camp violating its own health-safety rules. In March, hundreds of troops got sick and 38 died. And that was from what was likely a comparatively mild strain.
Conditions like those at Camp Funston were, of course, replicated all over the world, and connected by constant transport and movement during wartime. The flu followed, and eventually found its way into packed lodging homes and working-class apartments.
Sydenstricker published his first study of pandemic data on December 27, 1918 in Public Health Reports, “Preliminary Statistics of the Influenza Pandemic.” The first several paragraphs are devoted to his frustration with the existing data: “utter inadequacy and lack of uniformity of morbidity reporting… shortcomings of morbidity statistics… serious obstacles… a true prevalence rate can not be computed… comparisons of localities for prevalence obviously are impossible… the chronology of the epidemic in perhaps the great majority of localities has not been recorded, and if recorded is subject to serious errors… fatality rates manifestly can not be ascertained…”
To deal with the lack of data, Sydenstricker teamed up with Wade Hampton Frost, another PHS surgeon and the future head of the Johns Hopkins School of Public Health, who made the finding that the 1918 flu targeted young adults over children and the elderly, and who would later develop the concept of the index case. They hit the streets, reaching more than 46,000 people across Maryland from late November through mid December, and published a preliminary study by March 1919. Eventually the PHS surveys would reach nearly 150,000 people in 18 localities across the country.
Sydenstricker continued to report out the numbers after the pandemic faded away: a review of worldwide data, difficulties in reporting death rates, variations in case fatality. But he kept circling back to income and health care. In the thick of the pandemic in April 1919, Warren and Sydenstricker again collaborated on the Public Health Reports piece, “Health Insurance, the Medical Profession, and the Public Health.” The authors laid out the case again for a public health insurance safety net, comparing American outcomes unfavorably to the pioneering German system and the newer British system.
In 1920 Sydenstricker co-authored “A Method of Classifying Families According to Incomes in Studies of Disease Prevalence,” critical of the tendency in public health to put economic brackets in descriptive terms like “moderately well-off” or “well-to-do” rather than by income data. This was particularly important, Sydenstricker and his co-author wrote in the introduction, because “emphasis is being laid nowadays on the possible influence of environmental conditions” with the realization that the occurrence of most diseases “is intimately dependent… upon the mass of interrelated conditions under which a population lives.” He returned to his frustrations with descriptive income brackets in 1921, to tuberculosis in 1922, and published on nutrition, industrial-worker illness, income cycles, and worldwide disease prevalence in a busy 1924.
But all that was prologue to the Hagerstown morbidity studies. In 1921, Sydenstricker began a cohort study of between 8,000 and 9,000 people in Hagerstown, Maryland — more than a quarter of the town — and followed their health outcomes for more than two years, documenting nearly 20,000 illnesses. It would make up the basis for 11 separate studies by Sydenstricker from 1925 through 1929, and pioneer the field of long-term community health studies, defining its terms and cataloging its challenges.
The morbidity studies did not cover Hagerstown’s Black population, however. Sydenstricker’s explanation was that, at 5% of the population, “the number of negroes was too small to yield comparable results.” The late medical historian Harry M. Marks has documented how Sydenstricker’s work on pellagra, while a “radical social analysis” about the Southern cotton economy backed up with sophisticated new research methods, had a blind spot about race. As with the Hagerstown study, Sydenstricker and his co-author wrote that the Black population of the South Carolina mill villages they studied consisted of “few negroes” who “lived somewhat apart” and that it would be “disproportionately laborious… to secure all the desired data from the few negro families there.”
Why did Sydenstricker avoid the subject, figuratively and literally? Marks concludes that, because Sydenstricker came to health economics through industrial economics at a time when Black employment in industry was minimal, “African Americans had never registered with him as a significant group. As tenant farmers in the underdeveloped South, they simply seemed enmeshed in the same coercive labor system as poor whites… Race was a sociocultural category, a marker of the acculturation and material conditions of different immigrant groups. Literally and figuratively, for Sydenstricker, the social location of African Americans was off the map.” That is, Sydenstricker saw class, not race.
In that instance, this considerable flaw in Sydenstricker’s work, like its findings, would be echoed over and over. But his early training in industrial economics often proved beneficial, particularly when the country’s greatest economic crisis occurred. Attuned to how ill health could impoverish a worker, leading to worse health, leading to deeper poverty, Sydentricker put an emphasis on the much more difficult subject of morbidity over better-documented mortality data prepared him for the Great Depression. The death rate famously declined during that period, but Sydenstricker demonstrated that “disabling illness” increased among the worst off.
In the midst of all this, Sydenstricker returned, 13 years later, to the pandemic on which he cut his teeth. Appropriate for a forgotten pandemic, it was perceived to have devastated rich and poor in equal measure. It wasn’t true, and Sydenstricker was primed to show it.
Revisiting the data his team had collected during the flu season of 1918–1919, uniform patterns emerged. Infection rates increased as Sydenstricker moved down the economic ladder, with a ratio of 1.3 to 1 between the “very poor” and the “well-to-do.” (Even Sydenstricker was stuck in the descriptive framework he wished to change).
Perhaps more interesting was the case fatality rate (CFR). Sydenstricker found that the CFR for the well-to-do and moderate was 1.5. For the poor, it was a nearly identical 1.7. For the very poor, it was 2.8, nearly twice that of the highest economic tiers. He was modest but sophisticated about the implications. “The specific conditions that may be involved probably are not only numerous but are so intertwined that even a very intensive investigation of a very much larger exposure could give only partial and incomplete answers to the epidemiological questions that present themselves,” he wrote.
Sydenstricker had a point. A century later public health experts are not just trying to answer epidemiological questions about current diseases but the questions Sydenstricker and his colleagues had about the 1918 pandemic. In 2016, a team of epidemiologists, using a dataset of nearly 8,000 influenza and pneumonia deaths in Chicago during the 1918 pandemic, were able to establish a level of correlation unavailable to Sydenstricker. They found that mortality increased by almost a third for every 10% increase in the illiteracy rate. While Chicago’s socioeconomic map has changed over a hundred years, it hasn’t changed that much; the high pandemic mortality on the southwest and far south sides could be overlaid on any number of health outcomes today.
Yes, especially Covid-19. The disease followed a familiar pattern in Chicago. The first 10,000 deaths in Illinois began with an early breakout in the city, racing through Black communities with high levels of comorbidity. From there the pattern of deaths shifted to Chicago’s Latino population, which is younger but with high levels of essential workers in often crowded living and working conditions. The south and west sides have suffered high mortality rates from Covid-19, but when vaccinations started rolling out, they reached whiter, wealthier parts of the city first, just as Sydenstriker had found greater access to medical care in his groundbreaking Hagerstown studies.
Similar patterns play out nationwide. Black communities have the highest mortality rates from Covid-19. Younger-skewing Latino communities have disproportionately high case rates in many states, and as Sydenstricker spent much of his career emphasizing, morbidity is a critical measure despite rarely being the headline.
After his stint as a public servant, Sydenstricker became director of research at the Milbank Memorial Fund, a still-extant health policy think tank, while continuing on as a very active consultant to the PHS, where he continued to push for a greater role of the government in health care. For the Hoover administration’s Committee on Social Trends, Sydenstricker wrote a monograph, Health and Environment, that has been described as both “the first major treatise on social medicine” as well as “close to outlining an ecological approach to public health.”
Under Hoover it went nowhere. Sydenstricker worked on the Committee on the Costs of Medical Care (CCMC), but he wouldn’t sign the ultimately wishy-washy product. The committee’s final report broke significant ground on thoroughly documenting the problem but, in the end, would mostly precipitate a huge, ugly fight within itself and the medical establishment at large; the infamously “thuggish” editor of the Journal of the American Medical Association called it “incitement to revolution.”
After the failure of the CCMC, Sydenstricker began work with Isidore Falk, a former student and the research head of the CCMC, on a book about health insurance. They were interrupted with seemingly good news, pulled into the Council on Economic Security (CES) to design health insurance provisions for the Social Security Act. But the CCMC fight had poisoned the well.
“The political realities were such that all those responsible for trying to get the program through the Congress did not want to jeopardize the rest of the program by having an almost certainly futile fight in behalf of any health insurance measures,” wrote Thomas H. Eliot, the CES’s general council. “The fear was that if health insurance in any form was a part of the Social Security bill, this would arouse such vehement opposition that the whole bill would be killed.” Falk eventually became the main target for that vehemence, though he would persist through a career in government service, and, three decades later, eventually got to implement some of their ideas from the CCMC in a community-based HMO in New Haven.
Sydenstricker pleaded in Social Service Review that the “sensible co-ordination of public health functions with private medical practice… in no way postulates “any particular form of government,” much less “state medicine” or “‘Sovietized’ control.” Sydenstricker’s plans got sidelined again, this time under the best-case scenario of the Roosevelt administration and the most aggressive expansion of the social safety net in the country’s history.
Sydenstricker died the same year his piece was published. His diagnosis keeps circling back.