The Influenza Epidemic of 1918-1920 among the Navajos

Benjamin R Brady & Howard M Bahr. American Indian Quarterly. Volume 38, Issue 4, Fall 2014.

The influenza pandemic of 1918-19 was “the most widespread disease event in human history.” It “killed more people in a year than the Black Death of the Middle Ages killed in a century.” “Nothing else—no infection, no war, no famine—has ever killed so many in as short a period.” Perhaps one-fourth of the earths 1.8 billion people were infected, and over 50 million (2.8 percent) died from the disease and its complications.

Some populations suffered higher mortality than others. Across the globe, indigenous peoples were much at risk; remote indigenous populations were especially vulnerable. Many Native Americans, Pacific Islanders, and Africans—tribal peoples having more or less “naive” immune systems—”were not just decimated but sometimes annihilated” by the influenza. Native Americans “suffered hideously,” with mortality rates four times higher than in the wider population.

Studies on mortality differentials in the 1918 pandemic have identified several “associated factors” or “risk factors.” Taken together, these characteristics comprise a framework within which the experience of particular populations may be interpreted. In this article, the experience of the 1918 influenza among the Navajos is organized according to these factors gleaned from studies of the pandemic among other indigenous populations. We proceed from a review of these risk factors to a description of the 1918 influenza on the Navajo reservation, drawing on both Navajo and non-Navajo sources. We conclude with a reassessment of influenza-related mortality on the reservation, adjusting early-published official figures to reflect later data, some of it previously unpublished primary observation.


Our method may be summarized in four stages. First, we conducted a review of relevant historical and analytical literature on the 1918 influenza pandemic, combining works identified in contemporary databases (notably Social Sciences Citation Index / Web of Science) and library catalogs with an exhaustive “snowball” bibliographic search, tracking down items cited in works already identified, then following up relevant items cited in those works, and so on. We paid special attention to reports of unusually high influenza mortality and morbidity, noting population characteristics associated with high mortality. Second, we made a focused search for influenza-related materials in the Navajo literature, using the Internet and bibliographies specific to the Navajos and searches within the relevant time span of published and archival materials at the National Archives and in several southwestern university collections, including the libraries of the University of Arizona, Northern Arizona University, and the University of New Mexico. Third, drawing on materials identified in stage 1, we established a set of “risk factors” associated with high mortality during the 1918 influenza; these characteristics then served as an organizing framework for a summary description of the Navajo experience of the influenza epidemic, as revealed in stage 2. Finally, our consideration of primary and secondary accounts of the epidemic on the Navajo reservation makes possible a reevaluation of official reports on influenza mortality on the reservation, supporting the conclusion that official Bureau of Indian Affairs figures on influenza mortality among the Navajos in the epidemic are substantial underestimates.

Indigenous Populations and the 1918 Influenza

Relevant to interpreting the influenzas impact on the Navajos is the experience of other populations experiencing high mortality. The 1918 influenza was a new disease for which neither scientists nor medical personnel had a cure. Many of the responses recommended by physicians and other modernist experts at the time were no more effective than traditional remedies or techniques formulated in desperation by tribal peoples.

Early commentary on the 1918 influenza distinguished it from other diseases in that it seemed “socially neutral,” infecting people across nationalities, ethnic groups, and social strata. It later emerged that while the virulence of the disease was such that “everyone” caught it (much of the population was susceptible to this powerful pathogen), there were substantial social differentials in its lethality. In Britain, for example, 2.5 percent of the total population died, although almost 30 percent was infected. The same ratio applied to the United States, where around 28 percent of the population was infected and about 2.5 percent died. Embedded within these aggregate figures was substantial variation by location and ethnicity in both susceptibility to the disease and mortality from it.

In the Pacific Islands, Samoans had the most severe experience with the influenza. In a few weeks, about 8,500 died, a 22-25 percent population loss, all on the island of Western Samoa. On other Pacific Islands, at least 5 percent of the population succumbed. In the Caribbean, the influenza passed through the islands with extreme variation and selectivity. Some populations had high morbidity and mortality, while others remained virtually unaffected.

Among Native Americans there was great variation by tribe. According to case and mortality statistics compiled by the US Public Health Service in 1919, less than 1 percent of the Native Americans of Oklahoma, Wyoming, Kansas, and Michigan died from the influenza, compared to 4-6 percent in Arizona, Colorado, Mississippi, New Mexico, and Utah. Among Alaskan Natives, entire communities were stricken, and some towns were abandoned. In the vicinity of the town of Marshal, 30 percent of males and 10 percent of females died.

What was it about indigenous populations that made them especially vulnerable to the 1918 influenza? The literature suggests several factors, including (1) demographic characteristics, especially age and gender; (2) socioeconomic status; (3) the availability of epidemic-appropriate resources (nursing, food, shelter); (4) differential immunity related to health status and prior disease experience; (5) variations in social distancing, from overcrowding through quarantines; and (6) community organization and communication infrastructure.

For most affected populations, the 1918 influenza produced substantial written records. Drawing upon those data, researchers have identified population characteristics associated with high mortality. Table 1 summarizes the relation between these “risk factors” and mortality from the 1918 influenza. We consider each of the six risk factors separately.

Demographic Factors: Age and Gender

The 1918 influenza traced the world in three waves. The first, in the spring and early summer of 1918, was relatively mild but was closely followed by a deadly second wave that lasted from late August through early 1919. The third wave, “a relatively minor echo of the second,” followed in late winter and early spring of 1919. The influenza’s spread was hastened by its brief incubation period, twenty-four to seventy-two hours. The speed with which the disease spread was one of the reasons it was so lethal.

Many observers supposed it attacked indiscriminately, without regard to social status or gender, but it was soon apparent that age was relevant. Most influenza outbreaks killed the very young and the old; the 1918 influenza was atypical, and especially frightening, because it also killed healthy young people. Almost half of the its victims were adults in their twenties and thirties.

As summarized in table 1, most accounts treating gender mention higher mortality among men, although this pattern did not appear everywhere. Pregnant women were especially vulnerable; they were “those most likely of the most likely to die.” Among gender-related effects of the pandemic were that women were more likely than men to be nurses and caretakers, and thus the loss of a woman to the influenza might be more devastating to a family’s future than the loss of a man.

Socioeconomic Status

The flu threatened everyone but did not kill in equal measure. As illustrated in panel 2 of table 1, the poor were more likely to contract the flu and more likely to die from it. It was not wealth itself that mattered but whether the sick had help and resources that facilitated survival. Among other things, the well-to-do tended to be healthier, were better traveled and thus had had more prior exposure to illness, had better medical care, had more assistance to care for them, were better able to take time off from work to convalesce, and were better fed, clothed, and sheltered. Still, the rich died along with the poor; care was essential.

Epidemic-Appropriate Resources

Mortality was highest among groups collectively too sick to care for themselves. Studies cited in panel 3 of table 1 argue that surviving the influenza depended on the availability of nursing care, food, and shelter. The doctors of 1918 did not know how to cure the disease, but it did appear that nursing, or simple supportive care, helped. Study after study identifies good nursing as the most effective treatment at the time. Nurses, family members, and neighbors saved patients who otherwise would have starved or died from exposure because the flu had devastated their ability to perform even minor tasks.

Health Status and Immunological Experience

Immunity conferred by prior exposure probably meant the first wave contributed to lower mortality of the general population in Europe and America. Geographically marginal groups that first encountered the influenza in its deadly second wave were especially vulnerable, as were populations weakened by prior or chronic disease or having compromised health due to malnutrition or a history of poor medical care. Studies cited in panel 4 of table 1 suggest that in the face of an epidemic, risk may derive from physical isolation, physical debility, or any characteristic related to prior exposure to infectious disease.

Social Distancing: Crowding and Quarantines

Other things being equal, distance protected. The influenza spread through interpersonal contact; crowding, whether temporary at social events or more permanent, as in crowded lodgings or military camps, facilitated infection. The “distance” that mattered was separation from carriers of the influenza, and so quarantine, if total, was effective.

Both geographic and social qualities are involved in successful quarantine. A community must have organized leadership to coordinate containment efforts. Geographic factors may either facilitate or complicate the process. Distance may also work against the community: the more remote the area, the less controlled its borders tend to be and the harder it is to provide assistance if the quarantine fails. For containment to be successful, there had to be conscious, intended effort at enforcement. Only a few isolated communities were able to maintain effective quarantines, at least through some waves of the epidemic.

Community Organization and Communication Infrastructure

Influenza, like all communicable diseases, spreads socially; responses to it, to be effective, must also be social in nature. Because carriers were most infectious before they showed any signs of illness, the flu virus had the upper hand even before its presence was known. Findings illustrated in panel 6 of table 1 suggest that a centrally organized authority was useful in disseminating information and mandating compliance to “proper” response to the disease. However, compliance was often compromised because people had limited understanding and too little warning of what they faced.

Towns, states, and nations lacking the means and organized structure to promptly articulate and communicate the nature of the disease and appropriate responses were left to respond on their own, based on local understandings of the disease. The best most people could do was to follow commonly prescribed symptom-relieving protocols. Such practices helped reduce mortality in areas accustomed to treating the flu. In areas not accustomed, traditional healing practices sometimes amplified susceptibility to the disease.

The Influenza Epidemic of 1918-1919 among the Navajos

To judge from the published literature, the epidemic of 1918 is mostly forgotten, both by Navajos and by the general public. This is remarkable, as the epidemic was among the most fateful events in tribal history. Certainly, excepting the Tong Walk and the Navajo incarceration at Fort Sumner, it was the most costly in lives. Despite its extraordinary demographic impact among the Navajos—estimates of total mortality from at least three waves of influenza over this period range as high as 10 percent—several factors have combined to minimize remembrance, or at least the writing and talking about individual victims and the scale of loss.

Data Limitations

Most writing on the influenza among the Navajos draws upon the same few sources. Among the most frequently cited primary sources—eyewitness accounts from people living on the reservation during the epidemic—are two articles by Albert B. Reagan, a newly appointed Indian Service school administrator at the Marsh Pass school near Kay-enta who arrived in Navajo country in early October, just as influenza victims were beginning to die. Reagan helped nurse the sick in Tuba City and Kayenta and afterward traveled about trying to tally the number of deaths. Other firsthand accounts derive from the correspondence of agency physicians, memoirs by traders, and missionary writings. We might expect statements by the Navajos themselves, but accounts of their experience are rare. Because traditional Navajo culture defined the dying person as a threat to the living and discouraged talk about those who had died, and also because most Navajos did not speak English, there are few published reports on the epidemic by Navajos. Existing Navajo accounts typically were recorded by anthropologists or traders. The best published report on the epidemic by a Navajo is Rose Mitchell’s chapter, “The Flu,” told to anthropologist Charlotte J. Frisbee. Among the most cited secondary sources are agency reports to the Indian Bureau, accounts by reporters for local newspapers, and scholarly articles.

Even the best histories of the Navajos barely mention the influenza epidemic, or else they summarize it in a paragraph or two. Peter Iversons The Navajo Nation begins discussion of Navajo health care with the stock reduction era (1930s) and does not treat the earlier influenza epidemic. His authoritative Dine: A History of the Navajos subsumes the epidemic in a three-page section, “Tuberculosis, Trachoma, and the Flu,” devoting only one page to the 1918 epidemic. Garrick Bailey and Roberta Glenn Baileys well-reviewed A History of the Navajos: The Reservation Years devotes just over a page to the epidemic. Drawing upon agency superintendents’ reports, they offer a summary table showing mortality rates among the Navajo agencies ranging from 3 percent to 18 percent.

Wade Daviess Healing Ways, mostly devoted to post-1940 health care, has just two sentences on the 1918 epidemic. Davies references another medical history, Robert Trennert’s White Man’s Medicine, which gives four pages to the influenza epidemic. Trennert concludes that between 2,100 and 2,500 Navajos succumbed to the influenza between October and December 1918 for a mortality rate in the 7-8 percent range.

Factors Associated with High Influenza Mortality

Franc Johnson Newcomb described the rapid spread of the influenza: “The first persons to die were buried in the accustomed manner, but soon death struck too fast and the living members of the family were too sick and too weak to attend to the burials.” Mary Kennedy spoke of the “great force” of the epidemic, when “most of the tribe was laid low.” Hilda Faunce also described the rapidity of the influenza’s spread: “Like a grass fire the disease swept the Indian country. Every day some one told of deaths.” Rose Mitchell said that it started spreading “almost overnight.” Reagan tells of “a family of eight [who] were picking pinon nuts when the disease overtook them” and of a father and five children driving their sheep “when they were overtaken by the disease and died one by one along the trail.”

The 1918 influenza was remarkable in that it reached far back into the trackless areas of the reservation, where isolated families were uniquely vulnerable. Louisa Wetherill remembered that, in the spring of 1919, “riding out to remote hogans, the Wetherills found the bodies of the dead,” and that at the “height of Zilhlejini,” a mountaintop area, Ben Wetherill “found a hogan with five bodies in it. He closed up the doorway and left them there with the hogan as their tomb.”

Demographic Factors: Age and Gender

There is insufficient data on the number of victims, let alone their age and gender, to be certain whether mortality from the 1918 influenza on the reservation followed demographic patterns observed elsewhere. There are only hints, from observers’ impressions and from partial counts of the characteristics of victims. One trader wrote that “small children and old people were the first victims, but the flu played no favorites and soon the death rate was just as high among the strong men and women.” Tall Woman, who almost succumbed herself, remembered that the flu “seemed to really hit the young children; little children all over the area were dying day and night, night after night.” Illustrative numbers are available from Reagan, who counted the dead within a twenty-five-mile radius of the Marsh Pass Boarding School (Kayenta) and noted their age and sex. Government farmer Alva Shinn did the same for the area around Fort Defiance. Russell analyzed Reagan’s numbers and concluded that in the Kayenta area, “the influenza particularly affected infants, small children, and teen-agers.”

Our own analysis of Reagan’s data, including deaths among “children” whose age was not known and who therefore were not included in Russell’s count, indicates that about 60 percent of all deaths were of children below age fifteen. Shinn’s tally also suggests that over half of all deaths were children. Both tallies indicate that female children and older women were perhaps more likely to die than their male counterparts. The number of deaths among men and women in early and middle adulthood were roughly equal. Russell provides figures for Navajo population by age. In 1915 children under fifteen made up about 51 percent of the Navajo population, yet they were over 60 percent of the influenza deaths recorded in Reagan’s tally and about 52 percent in Shinn’s. As for gender, Shinn’s tally shows adult women more vulnerable than men (55 percent of adult deaths were women); Reagan’s shows female deaths among children almost twice as frequent as male deaths.

Socioeconomic Status

The Navajos of 1918 were among the poorest of the poor. Their destitution is apparent in statistics on income and property published a few years after the epidemic in the influential “Meriam Report.” It showed American Natives generally to be poor and that “several tribes, embracing in the aggregate a high proportion of the total number of Indians under government supervision may even be said to be extremely poor.” The report included statistics on annual per capita individual income and annual per capita individually owned property by jurisdiction. Among sixty-five Indian jurisdictions, the five Navajo agencies were consistently at or near the bottom, ranking 45th, 47th, 53rd, 57th, and 65th in individually owned property and 36th, 37th, 44th, 59th, and 64th in per capita individual income.

We recognize that defining poverty across cultures is problematic and that the Meriam Report’s use of individually owned property and individual income as indicators may miss, for example, collective or family resources. That the Navajos were “poor” in health-relevant resources is apparent in Trennert’s review of health care on the reservation in the early twentieth century. His history describes “government parsimony,” “effective delivery of medical services stymied at every turn,” “persistent complaints about the lack of medical resources,” “hospitals inadequate and underfunded,” “postwar financial retrenchment,” and “government health officers obliged to make do with what they could beg, borrow, or steal,” ultimately producing “medical disasters [which] overwhelmed the reservation population.”

Epidemic-Appropriate Resources

Many Navajos died from the 1918 influenza because there was no one to care for them; abandoned family members died of starvation. Scott Russell argues that “the dispersed Navajo settlement pattern adversely affected the availability of care” and that seeking assistance, as opposed to fleeing into the hills, was related to survival: “The death rate was lower among those that sought help.” The Navajos were dispersed in “camps,” with many families living at some distance from other Navajos. During the epidemic all family members might catch the disease at the same time, and “since families resided alone, no one was available to tend the sick.” Their dispersion did not keep them from catching the disease but was sufficient to prevent them from obtaining assistance. That care made a difference is apparent not only in accounts of low death rates in boarding schools where most students were afflicted but also in Navajo accounts of family care. It seems to have mattered less what foods, medicines, or rituals were administered than that the patients were fed and kept warm. The size and desolation of the reservation and the physical isolation of some of the camps prevented many Navajos from receiving aid. Physicians, nurses, traders, Indian agents, and agency workers might stretch themselves to their limits, but they were too few and the problem was too immense for them to reach Navajos who were isolated in the camps or sick and fleeing deeper into the wilds in an attempt to escape contagion and the ghosts of the dead. Anglo helpers had their hands full caring for sick Navajos at the settlements, trading posts, schools, and hospitals. Writing in defense of Superintendent Peter Paquette, of the Southern (Fort Defiance) agency, who following the influenza epidemic was investigated by the Indian Bureau, Father Anselm Weber stressed the utter impossibility of caring for all the sick Navajos or even finding them, given the level of federal support in 1918:

They have also investigated the number of Navajo that died in the camps of the influenza! Mr. Paquette did all he could for the camp Indians, visiting them with the doctor, bringing them to the hospital in his auto, and so on. Of course, he could do nothing for those living very far from the agency. There is a limit to everything. Or did the Indian Office have spare doctors and nurses that would have been sent to these distant camps? What about our military camps? Why not investigate and suspend and decapitate everyone in authority whose “subjects” died in large numbers for want of attention?

It was “want of attention,” more than anything else, that drove Navajo mortality so high. Father Berard Haile noted the utter absence of appropriate care in his Lukachukai district: “During the entire epidemic nothing was done in the way of relief, but when it had well disappeared, statistics of the deaths were taken.” Historian Robert McPherson concludes that “deaths among the vast majority of Navajos living at large were never investigated.”

Health Status and Immunological Experience

As a population the Navajos were weakened by decades of inadequate medical care, frequent epidemics of childhood diseases, and widespread tuberculosis and trachoma (a highly contagious bacterial infection of the eye that may lead to blindness). Trennert labels the early decades of the twentieth century on the reservation the time of “a scourge on the land.” By 1910, he writes, “medical conditions in Navajo country bordered on chaos, with disease spreading at a quickening rate.” Almost one-third of the tribal population was afflicted with trachoma, and tuberculosis was everywhere, accounting for almost half of all reported medical problems. Just prior to the US entry into World War I, “tuberculosis and trachoma remained unchecked, hospitals were inadequate and underfunded, and employees’ living conditions remained primitive.” With the war things got worse, because doctors and nurses on the reservation volunteered for military service.

Social Distancing: Crowding and Quarantines

There was at least one successful quarantine among the Navajos. Anselm Weber, father superior at Saint Michael’s, just six miles from the school at Fort Defiance, where almost everyone was sick, imposed a quarantine on Saint Michael’s School when he learned of the outbreak at Fort Defiance.

At the time we had 250 students in our school at St. Michaels. I conducted a strict quarantine there and ordered that no one leave the school grounds, and that no one, neither red nor white, could visit the school. Because we missionaries were daily placed in danger of infection, a few days later I had the music room of our school set up as living quarters and “banished” Father Celestine Matz, OFM, the chaplain of our school, there. We can thank these precautions and the prayers of the Sisters [teachers at the school] that our school has so far been completely protected from the influenza.

A later wave of the influenza did affect students at Saint Michael’s School, but it was far less deadly than the fall 1918 version.

At least three characteristics of Navajo culture related to social distancing increased mortality from the influenza. First, their usual cultural response to illness was to call a medicine man and have a “sing,” a ceremony attended by family members and kindred of the sick person. Since influenza was extremely contagious, sings aimed at healing tended to spread the disease.

Second, the Navajo belief that people approaching death posed a threat to the living, combined with the belief that a place in which someone died was haunted by the malicious spirit of that person, had led to the practice of moving dying persons outside the home. This meant that sick people were moved from the shelter of the hogan out into the elements, precisely the wrong thing to do with patients needing warmth and care. If someone died in the hogan, its other residents, despite their own need for care and shelter, would abandon the dwelling, exposing themselves to the weather.

Third, their pattern of living in “camps” of a family or two dispersed across the reservation, often far removed from other camps or settlements, meant that when several members of a family became ill at the same time, their isolation worked against them; there was no one to care for them. In numerous instances, their isolation was such that their bodies were not discovered until months after their deaths.

In passing, it must be noted that all cultures had patterns of association that helped spread the disease. “In essence, Spanish flu was unstoppable; unless you shut down society altogether, any airborne disease always will be.” Despite the influenza, the US military continued to crowd servicemen into troop ships and army camps, patriotic parades in major cities went on as scheduled, schools and theaters remained open in some cities, and even amid efforts at isolation and shut-down, “if significant numbers of people continued to climb onto streetcars, continued to go to work, continued to go to the grocer,” the virus spread.

Community Organization and Communication Infrastructure

Some sense of the inefficiencies of the communication infrastructure may be seen in the tardy identification of the influenza by Indian Office employees, persons who should have been better at reading the warning signs. Reagan writes that he traveled from Washington DC through Flagstaff to Tuba City and that along the way the newspapers were “filled with accounts of the ravages of the Spanish ‘flu.'” At Flagstaff he “found the state normal school closed on account of an outbreak of the disease.” Yet later at Tuba City, when two agency employees were “not feeling very well,” no one had the imagination to suspect they suffered from the influenza. Similarly, Louisa Wetherill, traveling the reservation to solicit donations of sheep for the war effort, experienced early symptoms of the influenza but, not recognizing them, continued visiting the camps, becoming a likely “unwitting transmitter” of the disease. Indian Service personnel, who might have suspected that the influenza observed elsewhere had finally reached the reservation, simply did not make the connection in time. When they did, they were so busy dealing with the “prostrate” Navajos that any possibility of early warning was lost.

The failure of the informal Navajo communication system to warn of the disease maybe attributed both to the inadequacy of the Indian Service infrastructure and to the influenza’s rapid spread. As the epidemic progressed, officials were aware of the Navajos’ needs, but bureaucratic statements of good intent were worthless on the reservation, where it was often impossible to carry out instructions. In the Leupp subagency, “despite heroic efforts to quarantine the school, the disease ‘spread like wildfire.'” The agency superintendent writes, “I could not get medical attendance of any kind for the Indians.” Trennert adds, “With neither the government nor traditional medicine prepared for the devastation, the Navajo would suffer significantly.” Commenting on conditions on the eastern side of the reservation, he notes that in the Fort Defiance sub-agency, “even if all the medical personnel had remained healthy, the flu could not be contained and it quickly reached the outlying population.” At Pueblo Bonito, “government services quickly broke down,” and “very little could be done, with the temporary doctor limited to handing out aspirin and quinine tablets. Corpses were left where they lay and the unopened Shiprock hospital became a morgue.”

As it became clear that the influenza left its victims in weakened condition, after which pneumonia finished them off, the government issued instructions on preventing the development of pneumonia. Mostly, such instructions were meaningless, for “few opportunities existed in the Navajo country to implement these precautions.”

Successful Navajo Responses

Although there are few accounts of responses to the 1918 epidemic from the Navajo standpoint, the histories note that “not all the heroic effort was expended by government personnel. The Navajos themselves also worked to alleviate the epidemic.” McPherson writes that “the Navajo response to this catastrophe came in two forms—spiritual and physical.” Medicine men wore themselves out traveling across the reservation administering Blessing Way and Evil Way ceremonies. They may have carried the disease, and the close contact among ceremony participants may have spread it further, but “in the mind of the Navajos these healers saved many lives and performed a valuable service comparable to the work of the [white] doctors.”

Some Navajo responses to the influenza were effective, although most accounts of the 1918 influenza neglect that fact. In interpreting accounts of the epidemic on the reservation, it is essential to remember that Western medicine was unable to cure the influenza. Everywhere, good care helped people survive; but doctors and nurses died along with everyone else, and supportive care in the Navajo context could be as effective as in the American city.

We have three accounts of successful Navajo treatment of the influenza, two by traders’ wives and one by a Navajo survivor. Hilda Faunce tells of the emaciated Hosteen Tso, who visited their Black Mountain Trading Post sometime after the sickness had passed. His family often, he said, was the only family he knew where no one had died from the flu. When two of his boys were stricken, he had ridden about for two days trying to find a medicine man. He had gone to one, then “another, and another, many of them, but they were sick themselves or were singing the chants for others who had the sickness.” He returned home to find everyone sick. Despairing of finding a medicine man, he “prayed to several deneh gods that knew me” and concluded that he himself must be doctor for his family. He collected plants and cedar berries, boiled them with water in coffeepots, and administered doses to his wife and children. As he served them he sang healing songs, and when the medicine was gone, he gathered more berries and plants and repeated the process. “There were days,” he said, “when no one came to my hogan. I did not sleep but sang the prayers and gave the medicine until all of my family were well.” His neighbors learned of his success and asked for help, and so “I went to many hogans, many of them. The food was not always good, and I did not sleep much, so I am thin.” Faunce predicted that he would now become “a big medicine man,” but Hosteen Tso said no. His healing experience was “only for this time and because my prayer was right for me and the need great that the gods helped, but no more. I am no doctor.”

A second account is by Tall Woman (Rose Mitchell), the daughter of Man Who Shouts, who was a headman and a medicine man. It is based on her own experience and what she was told by family members after she recovered from the flu. As word came that a killing sickness had invaded the reservation, her father talked to other headmen and medicine people and made a plan. He decided his people could better survive the sickness if they lived close together and could help each other, so he urged them to move into proximity: “He told us to stay close together, to help one another, and to comfort one another.” He also gathered plants and made medicines, some to be drunk, some to be mixed with fat and rubbed on the body, and others to be applied like poultices. During the epidemic he butchered horses so the meat could be boiled and used as broth; the fat was mixed with herbs in a healing paste. He stressed that this disease could not be cured by Navajo ceremonies, for “this kind of sickness, this epidemic, had nothing to do with any of our ceremonies, not even the small ones.” He told the people the best thing they could do was to pray, and “he, himself, prayed night and day, and even at different times during the day, using his Blessingway prayers to protect his own children and all the others around us, and to get the flu to leave the Navajo reservation.” Tall Woman remembers, “My father was the only one who kept riding around, visiting the people living near us to check on them and to see what he could do to help them during that sickness. He helped whoever was in need of help, and kept using up his horses for that purpose.” He also helped to bury the dead. The result of his ministrations, she notes, was that her people were very fortunate during the flu, “because we lost only a few people. In other places, entire families passed away.”

A third successful Navajo response to the influenza was by Hosteen Klah, “noted medicine man, wealthy stockman, and unsurpassed weaver” who lived near the Newcomb trading post. Klah returned from Colorado in the late fall of 1918, weak and fevered. He refused to stay at the trading post, where he could be nursed, insisting that he be moved to his hogan, where he could access his herbs and powders. He identified the disease with an earlier epidemic that had killed many Navajos and refused to have visitors lest the infection spread. Klah remained in his hogan for ten days, breathing herb fumigants and drinking native quinine tea. Thus cured, Franc Newcomb writes, “he was immunized and ready to care for his clan members when the influenza epidemic finally hit us.” When it came, Klah was kept busy “taking care of the sick of his own clan and the families who lived nearby. As their hogans were clustered in groups, he did not have far to travel. He was the most successful doctor of any I heard about, losing only one member of his family.”

Revising the Toll: A Reassessment of Navajo Influenza Mortality

Just as resources for surviving the epidemic were problematic among the Navajo, so too were resources for making sense of and remembering it. Navajo culture discourages talk about the dead, and high mortality may be combined with rapid “forgetting.” A reassessment of the literature on the 1918 influenza, including some newly discovered correspondence by one of the government workers charged with making the original mortality estimates, suggests that the official reports substantially undercounted Navajo deaths. Our case for underreporting rests on three separate arguments. First, the official count was closed too soon. Figures submitted in December 1918 are presented as estimates despite evidence that the influenza continued well into 1919. Second, some of the numbers reported, especially for the Western Navajo, do not square with qualitative accounts of the progress of the influenza at the time nor with later checks on some of the figures using genealogical and other data. Third, a previously uncited eyewitness account specifically devoted to checking figures that went into the 1918 totals suggests an undercount of at least half again as much as the original count.

The Early Close

Some writers have reported the quick passing of the epidemic. Trennert says, “The epidemic ended almost as quickly as it began. By December the disease had passed.” Bailey and Bailey comment that “for a shortlived phenomenon, the influenza epidemic had devastating effects,” adding that “some new cases appeared as late as December 1918.” Reagan does not give a terminal date for his count of the dead, but he refers to events of the fall of 1918 and speaks of the influenza having run its course that fall and even of its “final abatement.” Yet there were many new cases in early 1919.

Directly after the appearance of the epidemic in the fall of 1918, when the most lethal wave of the epidemic had subsided but before the disease had run its course, the Indian Service had its agents submit figures for total deaths. These totals from that fall became the official influenza mortality statistics for the reservation, despite recurrences of influenza in succeeding months. By early January the numbers had been collected and the agency narrative reports submitted. The federal response to the perceived exaggeration of Navajo mortality in local newspapers was dated February 9,1919. Yet in some parts of the reservation the influenza continued to sicken and kill through the spring of 1919, and it returned in a fourth wave in early 1920.

Franc Newcomb’s recollections about the influenza in the winter and spring of 1919 are supported by other reports. Rose Mitchell remembered that while the disease and the deaths were worst in the fall of 1918, they continued for several months into the new year. The influenza “stayed around, here and there, into the next spring. So, some of the People who died from that died later, the next spring, when it seemed to come back. That happened to them, even though they survived it in the fall.” Louisa Wetherill remembered not merely a month or two in the fall but that the terror continued into the spring of 1919: “With the coming of spring the deaths on the reservation became rarer. The People began to breathe again after the long months of terror.” Mary Kennedy also observed the epidemic firsthand from a trading post, writing that it extended through “fall and winter of 1918-19,” being most virulent before the deep winter snowfall.

Traders’ recollections that the influenza continued into 1919 are backed up by another set of eyewitnesses whose contributions have been neglected in the influenza literature. Franciscan missionaries working among the People participated with them in reacting to the epidemic, working in hospitals, visiting the sick and dying, and burying the dead. Charlotte Frisbie uses the parish records at Our Lady of Fatima Church in Chinle to document influenza mortality in 1918-19, with the last recorded death on March 13,1919. The Franciscan fathers also documented the influenza’s persistence on the reservation into 1920. On February 20, 1920, Father Ludger Oldegeering wrote from Saint Michael’s that “the influenza has not spread as widely this year as in 1918. We have no sickness here, but at Fort Defiance there were some 50 cases and at present there are still a few ‘flu’ patients at the hospital.”

The Incongruity between the Official Counts and Estimates from Other Sources

Several eyewitness descriptions call into question the official death tolls. Reagan emphasizes the deadliness of the flu outside the settlements where care was available: “When the disease struck the Navajos they fled from the places where it appeared, often abandoning everything in their panic. The fatality of the disease was astounding. Whole families were wiped out.” Wetherill writes that Navajos came every day to the trading post asking for someone to bury their dead, that “all over the reservation smoke [was] rising from the hogans of the dead,” and that some large families died together as the epidemic took its “tragic toll.”

Relevant to the reported 6 percent mortality for the Southern or Fort Defiance agency is trader Faunce’s description of the epidemic. Her images hardly square with the reported death toll of only one in sixteen: “Like a grass fire the disease swept the Indian country. Every day some one told of deaths. A dwelling where a death occurred was always vacated immediately; for miles around every good winter hogan was deserted. The living moved out into the rain and found what shelter they could in temporary camps. Here death came again and those who were left moved on.”

Given the statements by eyewitnesses on the Western Navajo reservation of the “appalling” death rate, the “astounding” fatality of the disease, the smoke “rising from the hogans of the dead” all over the reservation, it is scarcely credible that the official report filed by the agency superintendent emphasized the general health of the Navajos and reported only two hundred deaths for the entire Western Navajo area. We noted that Reagan carefully documented deaths from influenza within a twenty- five- mile radius of the Marsh Pass (Kayenta) Boarding School. In that limited area, he recorded 148 deaths. It seems more reasonable to doubt the official tally than to believe that the fragment of the Western Navajo area Reagan visited accounted for three- fourths of all fatalities in the entire region.

There is also a striking disconnect between the reported two hundred deaths for the entire Western Navajo agency and later assessments of the effects of the influenza there. Scott Russell, who knows the Western Navajo area personally and whose grandmother was a victim of the epidemic, questions the positive 1918 official report on the Western Navajo influenza experience. His data suggest that the Navajos of the Western agency “were particularly affected by the influenza, though this is not revealed in [official statistics].” Russell even questions Reagan’s tally of 148 deaths in the Kayenta area because “from genealogies I have collected I know it to be incomplete.” He concludes that the official statistics on influenza mortality in the Western Navajo agency are wrong, that the consequences of the epidemic there were greatly underestimated, and that “in this area of the Navajo reservation perhaps 15% of the population may have died from influenza or its complications.”

Questioning the Original Counts: Alva Shinn’s Reliability Check

Alva Shinn, a long- term employee of the Indian Service, was government farmer at Lukachukai, Arizona. He was among those consulted in December 1918 by Indian Service personnel for estimates of Navajo deaths from the influenza epidemic in his jurisdiction in the Southern (Fort Defiance) agency. His 1918 estimate, made in “the excitement of the moment,” was a serious undercount, as he later would demonstrate. In October 1919, he wrote to Arizona congressman Carl Hayden, enclosing a report listing who had died in his district, thereby explicitly adjusting his original estimate of influenza deaths upward by more than half. Shinn wrote:

If I recollect, my report last December to the Department, was 100 deaths at Lukachukai District and may reach 150. It was truly made under the excitement of the moment, but as a statistical reporter for years for Department of Agriculture, my reports are very conservatively made, and my report on Influenza deaths last fall from Lukachukai District was not guessed at, but carefully based on reports made me, by my reporters, and I am sorry to say the facts revealed exceed my report.

Bothered that his December 1918 report was inaccurate, Shinn set himself the task of documenting precisely how many Navajos had died in his district and using that information to lobby Congress for better medical care for the Navajos. He also gathered estimates from informants in adjoining districts but admitted that his tally of deaths elsewhere was “only partially complete and not sufficient to strike a fair percentage of loss.”

Shinn reported deaths within his district by six geographic subdivisions and by each victim’s family name, gender, and age category (man, woman, boy, girl). He described his method thusly:

This report was taken by me, from the living left of families, and by word, from nearest neighbors of families wiped out! It is not guessed at. I recorded the names of families in my death book record. It took time, care, and difficult work. With my own hand I buried many, left on top of bosom of mother earth, and in silence tears fell, from my eyes upon the decayed forms of these I knew. This report is as truthful as I can give, checked and rechecked. It was a sad work, and of en, I held my pencil, stopped recording, while tears fell from eyes of survivors of Navajo families who could not speak, while I told my interpreter to tarry a moment, in silence, till the wave of grief passed away! I hope this effort of mine, to get the truth, will be appreciated by the Department.

His letter to Congressman Hayden emphasized the marginal position of the Navajo as compared to citizens in urban places and lamented the minimal assistance available to Navajos during the epidemic:

During the last epidemic, no doctors or hospitals in certain localities, for many miles upon this Reservation. We worked day and night. Hundreds had no medical care. We could [have] saved most of the little children and babies, if we had hospitals to remove them too but today, many sleep under these pinon trees, at Wheatfield and Tselili and with my own hand buried some, in the spring of 1919.

Shinn collected names of the dead in five areas of his jurisdiction. For the sixth, Black Mountain, he wrote: “I have been unable to take the census of deaths from influenza, in this part of my district. Yet it was heavy. I will estimate at 40 or 50 at least.” In computing the total for his entire jurisdiction, he used the lower figure. He also admitted that his documented totals were incomplete, adding an estimated fifty victims whom “no man will ever record at least, can’t be found, in entire District, estimated low, as I feel to do so is [appropriate].”

If we count only those victims of the influenza whom Shinn was able to identify by name, the total for his district is 239, as compared to his “report last December to the Department [of] 100 deaths in Luka-chukai District and may reach 150.” In other words, the documented deaths in part of his district were 50 percent higher than his previous high estimate and over twice as high as his low estimate of December 1918. If we also add in Shinn’s “low” estimates of undocumented deaths, the total of 329 for his district is over twice as high as the high estimate officially recorded the previous December.

Shinn’s letter to Representative Hayden recommended that Hayden transmit the report to the commissioner of Indian Affairs, and Hayden did so. The Indian Bureau’s official response to Shinn’s voluntary effort to correct a mistaken count now masked in an official aggregate total was written by Assistant Commissioner E. B. Meritt, with no copy to Representative Hayden. Shinn had written, “I hope this effort of mine, to get the truth, will be appreciated by the Department.” Meritt’s response dashed that hope. Instead, Shinn was told to mind his own business.

While it is not intended to limit your interest in general health conditions, it is desired that your activities as an employee touching such matters shall cooperate with the Superintendent of the jurisdiction wherein you are assigned to duty and that any report for official use be forwarded through him. You are also advised that the Office, through a system long in use, collects statistical and other information directly from the various superintendents, and it is preferred that you do not undertake this work in a special way, but that your time and service be faithfully given to the duties of the position you occupy as farmer, which of course include all reasonable efforts for the promotion of health and better home life within your district.

Shinn’s “truth” apparently never went beyond Meritt’s office, and Indian Service estimates of Navajo mortality from influenza in the fall of 1918 were not corrected.

Shinn’s report is especially valuable for three reasons. First, it includes estimates of mortality rates for segments of the Northern or San Juan agency, the only agency where the superintendent was unable or unwilling to supply the Indian Bureau with mortality estimates for the fall of 1918. Using Shinn’s partial figures as a way to estimate a reasonable total for the Northern (San Juan) agency is beyond the scope of this article, but it might provide a more grounded estimate than the usual unsupported guess that the missing data necessitate. Second, descriptive statements in Shinn’s independent account of the effects of the epidemic corroborate some of the effects Reagan reported for the Western Navajo (Tuba City—Kayenta) agency. Finally, comparing Shinn’s careful, after- the- fact accounting to the totals passed up the line for his district in December 1918 gives us a rough estimate, for that district at least, of the extent of underreporting. Several contemporary observers claimed Navajo mortality to be much higher than officially reported. The Shinn report solidly documents that fact for the Lukachukai district.

Assuming Russell’s 15 percent mortality estimate for the Western Navajo, and the most conservative estimated correction for the Southern Navajo agency based on Shinn’s reassessment (1.5 times the official figures), we offer in table 2 corrected totals for Navajo deaths in the 1918 epidemic.


Why was Navajo mortality from the 1918 influenza epidemic so high? The Navajos represent an extreme case among Native Americans, a large tribe with unusually high losses to the influenza. Our review of the experience of marginal populations elsewhere—an integration of many particular accounts of high influenza mortality—points to several common factors. On all of these—demographic factors, low socioeconomic status, little access to epidemic- appropriate resources, poor health status and immunological experience due to preexisting medical problems and poor nutrition, social distancing that stranded the infected and exposed them to unhealthful conditions, and limited community organization and communication infrastructure—Navajos were marginal in ways that increased their vulnerability for infection and, if infected, reduced their chances of survival.

All these factors fall within the category “vulnerability,” defined in terms of distance, duration, communication, organization, definitions of the situation, and social and cultural resources. Populations that were multiply vulnerable had heightened risk to infection and death by influenza. In comparison with other marginal populations worldwide, the Navajos’ situation of 1918-19 was an almost “perfect storm,” in that they manifested the marginal extreme on almost every risk factor identified. Considering the intersections of types of vulnerability exhibited by the Navajos, it is remarkable not that so many of them were lost but that so many survived.

Alva Shinn’s careful effort to produce an accurate death record and correct his 1918 report, long buried in the Indian Bureau archives and not acknowledged publicly at the time, is a dramatic demonstration of serious undercounting of deaths in one part of the reservation. Shinn’s findings are congruent with the recent evidence that worldwide, the human toll of the Spanish influenza was substantially underestimated.

It may be asked why amending the picture of the Navajos’ influenza experience matters: What is gained by correcting the official reports? First, it is an indication of the Navajos’ resilience. The losses they experienced were greater than most of the rest of the nation, greater than many other tribes, greater than the government would acknowledge. The tribe accepted, seemed to forget, and moved on, perhaps, as Tiana Big-horse remembered, with some adjustments to their concept of warrior: “In Navajo, a warrior is the one that doesn’t get the flu when everyone else does—the only one walking around, making a fire for the sick, giving them medicine, feeding them food, making them strong to fight the flu.”

Second, the heavy losses were an indication of federal neglect in preventive care, communication, and provision of care during the epidemic. This neglect becomes apparent in federal efforts after the fact to resist upward adjustments to the official death toll and to blame local administrators in investigations that followed in agencies with unacceptably high mortality.

Finally, a richer, more accurate account of the Navajo experience of the influenza epidemic may be a solid step away from forgetting and toward remembering. Obviously it was not only the Navajos who “forgot” the 1918 influenza; almost everyone did. Worldwide, this “demographic catastrophe” is underrepresented in collective memory and in histories of the period. Crosby calls it “America’s forgotten pandemic,” and Iezzoni writes that “as soon as the horror vanished from the face of the earth, the forgetting began.” She characterizes the influenza of 1918 as “a story of dying and forgetting, of bewilderment, chaos, bravery, and horror. This is a story of people who worked together and a country which almost fell apart. Yet, this is a story no one knows and few remember.”

Among the Navajos, there certainly was bewilderment, chaos, and horror, but there also was bravery, love, sacrifice, and commitment. These exemplary experiences will only be remembered when the influenza epidemic of 1918 is accepted as part of tribal and national memory.

TABLE 1. Social characteristics associated with influenza mortality, 1918-1919

Characteristics Relationship Sources Illustrative Findings'”
1. Demographic factors: age and gender Mixed: atypically high mortality among young adults; most sources suggest higher mortality among men; others report somewhat higher mortality among women Barry 2005 Bristow 2010 Echeverri 2003 Noymer and Garenne 2003 Phillips and Killingray 2003a Taubenberger 2003 Taubenberger and Morens 2006 Tomkins 1992 Zylberman 2003 Hays 2005, 394: “Women generally suffered higher mortalities, perhaps because pregnant women were especially vulnerable, perhaps because women (as the primary caregivers themselves) were less likely to receive nursing care when they fell ill.” Johnson 2003,141: “While it was the young adults who suffered the greatest mortality above all others, there is little reporting of any significant variation in mortality by sex. Again, this is more or less consistent throughout the pandemic the world over.” McCracken and Curson 2003,121: “Overall the epidemic took a substantially heavier mortality toll of males than females.” Taubenberger and Morens 2010, 21: “The age group affected most severely by the 1918 pandemic was that between 20 and 40 years, accounting for almost half of influenza deaths.”
2. Socioeconomic status Inverse: mortality rates higher among the poor than for others Davies 2000 Johnson 2003 Kraut 2010 Mamelund 2006 McCracken and Curson 2003 Patterson and Pyle 1991 Phillips and Killingray 2003a Ramanna 2003 Zylberman 2003 Hays 2005, 387: “Within India Hindus of low caste suffered much higher mortalities than those of high caste, illustrating the vulnerabilities of poor populations everywhere.” Mamelund 2011, 56: “Low socioeconomic status … is known to be associated with poor nutrition, more crowding, a higher disease load, and … high Spanish flu mortality.” Sydenstricker 1931,160: “Fatality . .. among ‘poor’ and Very poor’ persons was higher than among the ‘well-to-do’ and those in ‘moderate’ circumstances.”
3. Epidemic-appropriate resources (good nursing and supportive care; little else worked) Inverse: appropriate care reduces mortality Barry 2005 Bristow 2003 Crosby 1989 Hays 2005 Herring and Sattenspiel 2003 Keeling 2009, 2010 Lautaret 1986 Patterson and Pyle 1991 Tomkins 1992 Echeverri 2003,187: “The difference between life and death could depend … on such simple factors as the care and feeding of the ill.” Iezonni 1999,120: “Given the lack of a pharmaceutical or therapeutic cure, nurses were arguably more valuable than doctors and even more scarce. It was nurses who performed the small tasks which save lives, providing a sip of water, a warm blanket, warm food.”
4- Health status and immunological experience Mixed: mortality reduced by prior influenza exposure, increased by debilitation from malnutrition and other diseases Crosby 1989 Davies 2000 Killingray 1994 Mamelund 2003 Quinn 2008 Schoenbaum 2003 Barry 2005, 363: “Populations whose immune systems were naive, whose immune systems had seen few if any influenza viruses of any kind, were not just decimated but sometimes annihilated.” Hays 2005, 386: “German soldiers whose resistance, depressed by poor diet, was far lower than that of the well-fed Allied troops” were “laid low” by influenza. Mamelund 2011, 59: “The adult age pattern of death from the 1918-20 pandemic may be explained by differential immunological memory influenced by the long-term place of residence.”
5. Social distancing: crowding and quarantines Mixed: greater distancing reduces contacts and slows disease spread, but isolation increases mortality; crowds and crowding facilitate infection Barry 2005 Davies 2000 Herda 2000 Herring and Sattenspiel 2003 Iezzoni 1999 Mamelund 2011 Killingray 1994, 61: “The virus is transmitted from person to person by the respiratory route…. Close crowding of people offers ideal conditions for infection.” Phillips and Killingray 2003a, 9: “The poor, and those living in overcrowded and insanitary conditions, were … more likely to catch and to die from the virus.” Quinn 2008,135: “Only in rare cases was isolation sufficient to keep a community free of disease—it seemed to attack almost every island on the globe.” Tomes, 2010, 49, 52: “Gathering bans, school closures, and other social-distancing measures significantly reduced mortality rates. … [Quarantine measures had to be accompanied by broader measures aimed at regulating the congestion of public spaces.”
6. Community organization and communication infrastructure Inverse: well-organized response reduces mortality Barry 2005 Crosby 1989 Davies 2000 Hays 2005 Iezzoni 1999 Jones 2002 Rice 2003 Aimone 2010, 78: “New York City ‘escaped’ with a low mortality rate because of the city’s health efforts over the previous 20 years…. [It applied] a variety of time-tested and adaptable regulatory and voluntary techniques already at its disposal.” Ramanna 2003, 95-96: “[In India] mortality would have been reduced had it been possible to provide immediate medical aid and suitable nourishment to those attacked.… The absolute lack of any public health organization in rural areas and the total helplessness of the rural population in the face of the calamity are evident.” Tomes 2010, 49: “Those municipalities that were able to quickly minimize the mixing of people in public spaces lost fewer lives to the Spanish influenza.”

Note: Details for all the sources mentioned in this table can be found in the works cited list that follows the endnotes of this article. For each characteristic, sources cited under “Illustrative Findings” are not duplicated in the “Sources” column.

Table 2. Estimated deaths, 1918 influenza epidemic, Navajo reservation, adjusted for official undercount

Agency 1918 Population Deaths (Official) Deaths (Corrected) Deaths (percent)
Leupp 1,411 144 144 10
Southern Navajo (Fort Defiance) 12,080 780 1,170a 10a
Western Navajo (Tuba City) 6,087 ca. 200 913[b] 15[b]
Eastern Navajo (Pueblo Bonito) 2,724 ca. 500 500 18
Northern Navajo (San Juan) 6,500 ? ca. 650[c] 10c
Total Navajo reservation 28,802 ca. 2,000 3,377 12

Source: Columns 1 and 2, and the overall format of the table, are from Bailey and Bailey, History of the Navajos, 119. A) Adjusted by a factor of 1.5, in line with Shinn’s documentation that the number of Navajo influenza victims in his district was, at minimum, half again as high as the total reported in December 1918. B) Adjusted to 15 percent of total population in line with Russell’s (“Navajo and the 1918 Influenza”) critique of the official estimate from December 1918. C) Estimated on the basis of Shinn’s partial count of deaths in some areas in the Northern Navajo agency. His counts suggest relatively high mortality (e.g., in the vicinity of one trading post, 30 percent of the infected Navajos died). Having no reason to believe that mortality was unusually low in the San Juan agency, we have set the estimated mortality at 10 percent, the lowest rate found among any of the other agencies.