The unprecedented ferocity of World War I (1914-1918) saw the great powers of Europe galvanize their populations into “total war” and the Continent besieged by violence, chaos, and destruction. Ultimately, the conflict resulted in the disintegration of several empires; witness the centrifugal fragmentation of Austria-Hungary, the dissolution of the Ottoman Empire, and the subjugation of Germany under the punitive Treaty of Versailles.1 The Great War, as it was also known, cut down a generation of young men in Europe and directly contributed to the emergence of a deadly influenza pandemic, known at the time as the Spanish Flu. The manifestations of pandemic influenza were horrific. Infected individuals exhibited heliotrope cyanosis, a bluish discoloration of the skin resulting from suffocation as their lungs filled with blood and fluid. The end result for the host was often systemic hemorrhaging, and ultimately death.

I begin this chapter by examining the impact of the 1918 pandemic on the institutions of affected sovereign states, their societies, and the interface between the two. A comparative demographic analysis of available mortality and morbidity data is followed by a description of the peculiar etiology of the 1918 pathogen. I then explore the effects of the pandemic on the state, and in particular its effect on the military forces (and societies) of the various protagonists in World War I, primarily through the available German, Austrian, British, French, and American data. Following that, I analyze the effects of the pandemic on governance (broadly construed) in affected polities.

Demographic Impact

During the final year of the Great War, pandemic influenza affected and debilitated circa 25 percent of the global population, typically resulting in the mortality of between 2 and 4 percent of those afflicted. Death typically resulted from influenza-induced hemorrhaging or suffocation, or from secondary pneumonic or tubercular co-infections.2 In the United States, flu-induced morbidity was circa 25 million, with an estimated mortality of 675,000. According to the medical historian Hans Zinsser, the epidemic exerted a dramatic (if rather brief) negative effect on average US life expectancy, resulting in a significant decline of 12 years in 1918.3

“The pandemic,” Alfred Crosby concludes,

affected history in general in the way that the random addition of a poison to some of the refreshments served at the 1918 West Point commencement celebrations would have affected the military history of World War II; i.e. it had enormous influence but one that utterly evades logical analysis and that has been completely ignored by all commentators on the past. On the level of organizations and institutions—the level of collectivities—the Spanish flu had little impact. It did not spur great changes in the structure and procedures of governments, armies, corporations, or universities. It had little influence on the course of political and military struggles because it usually affected all sides equally.4

Crosby’s argument is based on a reading of US mortality data; however, the data presented below indicate that the pandemic did not affect all the protagonists equally. The balance of evidence indicates that the virus generated differential mortality across the spectrum of affected societies. The very fact that mortality varied so greatly across cultures leads to the conclusion that the pandemic had differential impacts on the various combatants involved in the war. A second point is that there was considerable temporal variation in the waves of pandemic influenza that circulated the world in 1918-19, and that it struck and debilitated the Central Powers before it struck the Allies.

The first evidence to challenge Crosby’s assertion that all sides were affected equally by the pathogen comes from the medical historian W. H. Frost, who used the rates of mortality in mid-size to large population centers to document the degree to which influenza swept the United States in 1918. Importantly, Frost’s data clearly indicate differential rates of morbidity across US population centers, ranging from 15 percent in Louisville to 53 percent in San Antonio.5 Reinforcing this finding that flu-induced mortality was not uniform, but rather ranged along a continuum within societies, the medical historian Edgar Sydenstricker estimated that US national mortality rates ranged from 2.76 percent to 4.6 percent.6 Considering this estimate in terms of rates, Crosby noted that (according to US Public Health Service surveys conducted at the time) 280 per thousand US citizens contracted pandemic influenza in 1918-19. Crosby extrapolates to conclude that over 25 percent of the US population was infected and debilitated by the contagion.7

Given that influenza-induced morbidity and mortality appears to have ranged along a continuum within societies, one might expect to observe considerable variance between sovereign states. The data bear out this supposition, as certain countries (e.g., Japan) exhibited exceptionally low mortality rates, whereas other countries exhibited exceptional to catastrophic levels of mortality (the worst case being Samoa). The medical demographers G. Rice and E. Palmer analyzed Japanese medical archives to compile data on influenza-related morbidity and mortality, and determined that Japan witnessed 2,168,398 cases (morbidity) and 257,363 deaths (mortality). “The case rate,” they write, “was therefore 370 per thousand, or just over one-third of the whole population, which was rather higher than that of the United States. However, the influenzainduced crude death rate was rather minute, at 4.5 per thousand.”8 Other societies were not so fortunate. Data collected by Colin Brown indicate that Indonesia’s mortality rate was approximately 17.7 per thousand.9 Furthermore, approximately 3 percent of Sierra Leone’s indigenous population died as a direct result of influenza by late 1918,10 and Patterson has established that flu-induced mortality in African societies ranged from 30 to 40 per thousand.11

Attempts to quarantine Australia and New Zealand were partially successful; they only delayed the onset of the contagion, and Rice notes that New Zealand’s Maori population exhibited a mortality rate of 43 per thousand.12 Mills has established that India suffered to an even greater extent from the virus, with a mortality rate ranging from 46 to 67 per thousand, again varying by region.13 The highest death rates appear to have occurred among the isolated and immunologically naive populations of islands such as Western Samoa, which exhibited a staggering mortality rate of 220 per thousand, resulting in the destruction of over 20 percent of its population base over the duration of the pandemic.14 (See figure 3.1.)

Initial estimates of deaths induced by pandemic influenza placed aggregate global mortality at circa 21 million. However, recent epidemiological investigations have revealed that flu-induced mortality in South Asia alone (particularly in India) exceeded 17 million. Therefore, conservative revised estimates of mortality currently approach 50 million,15 and liberal estimates are as high as circa 100 million.16 For the purposes of this inquiry it seems prudent to adopt the figure of 50 million.

During typical manifestations of the pathogen, influenza is a killer of those at the two tails of the demographic distribution of a society: the very young and the elderly. Yet the 1918 epidemic displayed an unusual penchant for the destruction of healthy and productive individuals in the prime of their lives. Specifically, during the 1918 pandemic, the mortality distribution associated with infection exhibited the form of a W, with pronounced mortality in the 15-45-year age range, accompanied by the expected high mortality in the elderly and young.17 Flu-induced mortality seems to have affected females and males in equal fashion, although the pathogen apparently generated exceptional mortality in pregnant mothers.18 Why would the pathogen affect so many healthy young adults in the prime of their lives? It is reasonable to speculate that the influenza generated a profound overreaction by the body’s immune system, and that the cytokines (endogenous toxins) released by the body destroyed the fragile tissues of the lungs during the immune system’s attempt to combat the virus.19 It would seem, then, that those with stronger immune systems were, as a perverse consequence, more vulnerable to the pathogen.

Beyond the influenza, various pathogens exhibited a pronounced and deleterious effect on the German population during World War I. One reasonable explanation for such declines in German public health is that the embargo on the shipment of foodstuffs to the Central Powers during this period would have severely compromised the base health of the average German citizen, increasing the probability of colonization of the human host by the pathogen. Indeed, male civilian deaths in Germany peaked in 1918 at 566,077, with female mortality in the same year reaching a zenith of 644,163 even though females were non-combatants. Compare such figures with postwar baseline civilian mortality of 429,741 for males and 426,263 for females in the year 1923.20 Note that this post-conflict baseline may be rather inflated relative to prewar data, owing to the fact that the war generated attenuated negative impacts on human health, ranging from immunosuppression and secondary infection to mental illness.21

Etiology

Recent evidence suggests that the 1918 influenza was in fact an H1N1 variant, and therefore genetically similar to the virus currently spreading throughout avian populations in East and Southeast Asia. The 1918 virus likely originated in avian species, crossed over into the human ecology through processes of zoonotic transmission, then continued to evolve and mutate within human populations. This helps to account for the three waves of the pandemic that circled the world in 1918, each progressively more lethal, and likely intensified by the conditions of World War I.

The orthodox epidemiological history traces the origins of the pandemic to Camp Funston (near Fort Riley, in Kansas) during March 1918, after which it appeared at Camp Oglethorpe in Georgia and then at Camp Devens in Massachusetts.22 With troop transport vessels serving as vectors of both incubation and distribution, the flu then supposedly traveled to the battlefields of Europe, whereupon it infected thousands of soldiers during the spring of 1918. During this initial phase, the pathogen exhibited significant morbidity, with slightly elevated mortality, and then entered a period of relative dormancy during the summer months that followed. Late August of 1918 saw the second wave of the epidemic erupt with much greater lethality, appearing simultaneously in France, Sierra Leone, and the United States (with its epicenter in Boston).

The third and most virulent wave of the epidemic appeared in the fall of 1918, attacking the military forces of both the United States and the Central Powers and overwhelming field hospitals, transports, and lazarets in the rear with fevered, debilitated, and dying young men. According to the microbiologist Paul Ewald, throughout these three increasingly lethal waves of infection, the influenza virus progressively mutated to take advantage of the densely packed populations and of the mobility of soldiers, resulting in the emergence of a highly communicable and lethal strain. Ewald notes that the mobility of forces acted as a “cultural vector” to distribute the virus from infected host populations to uninfected populations:

Soldiers in the trenches were grouped so closely that even immobile infecteds could transmit pathogens. When a soldier was too sick to fight, he was typically removed by his trenchmates. But by that time trenchmates often would have been infected.23

Thus, the malign ecological conditions associated with a protracted ground war allowed the virus to mutate in order to become more infective (and increasingly lethal) to the young adult populations that served as hosts in the theater of war, and resulting in the unusual W-shaped mortality distribution. Ewald writes:

The increased mortality in the trenches due to fighting or the other infectious diseases that typically accompanied such warfare should have, if anything, also favored a high level of virulence. Any deaths of recovered immune individuals would result in the transport of replacements into the trenches who would often be susceptible to the strains circulating in the trenches. In addition, one of the costs that a pathogen may incur from extremely rapid reproduction is a shortened duration of infection due either to a more rapid immune response or to host death.24

Thus, the epidemic was a product of the pernicious ecology of war, with high population densities of combatants, poor sanitation, stress, and the movement of forces collectively serving as remarkably efficient vectors of transmission around the world. However, the pandemic may have also affected the course of the war, to some degree, through its successive waves of debilitation and destruction of human life. Thus, we may understand the relationship between war, pathogenic emergence, and outcome of the war as a complex feedback mechanism. I shall explore this concept in greater detail below.

On the Political Suppression of Data

The effects of the contagion were historically downplayed by the medical community, who (like the Galenists of old) were acutely embarrassed by their impotence in the face of such an overwhelming epidemic. According to the medical historian Carol Byerly, “the tendency of medical officers, army commanders, and federal officials to downplay the role of the influenza epidemic in the Great War, and the impact of disease on military populations in general, has encouraged American complacency about the ability of medicine to control disease outbreaks during war.”25 The German medical historian Wilfried Witte has noted that Prussian authorities went out of their way to downplay the severity of the pandemic during wartime, going so far as to repress the dissemination of data as best they could. Furthermore, existing data suggest that, while the majority of deaths were recorded in urban centers, there was significant debilitation and mortality in rural areas that likely went unrecorded because of a lack of medical personnel in those regions. Therefore, many of the mortality data from Central Europe are probably significantly lower than the actual mortality that occurred as a result of both repression and low levels of health-care capacity in rural regions.26 Johnson and Mueller concur with the assessment that available estimates vastly understate the impact of the contagion on mortality. They posit that “limitations of these data can include nonregistration, missing records, misdiagnosis, and nonmedical certification, and may also vary greatly between locations.”27

The general dearth of investigation regarding the political and economic effects of the Spanish Flu are notable, and necessarily raise some questions. “The influenza epidemic’s most important, if enigmatic, legacy,” Byerly argues, “has been its reinforcement of the government’s and the society’s reluctance to acknowledge the deadly role disease often plays in war. As people have written war stories and official reports of wars, they have often effaced human suffering, reflecting the military’s tendency to downplay the fact of injury as a product of war. This tendency is especially apparent with respect to the story of disease in war.”28

Because death from disease was often regarded as ignoble, in the aftermath of the Great War military forces sought to diminish the perceived impact of the contagion, particularly since military medical officers had been utterly powerless in the face of such a pathogen. Byerly castigates the US government for its deliberate and widespread attempts to suppress evidence of mortality from disease during the conflict. She argues that the War Department’s official record of the war “listed battle casualties only” and that “a 1919 Senate document on the cost of the war stated that 50,000 men were killed in battle. Although this report calculated various costs of the war . . . it failed to mention the war cost of 57,000 deaths from disease. It reduced the army death toll by more than half.”29

Effects on the State

Central to this analysis is the effect of pathogens on the apparatus of governance within a polity, particularly the bureaucracy and the military.

The profound morbidity and mortality induced by the pandemic affected the armed forces of the various combatants in different fashions. Much as countries exhibited differential mortality from the contagion, the armed forces of states exhibited varying death rates. Comparative analyses of the causes of death for military forces, measuring mortality from disease versus that suffered in action, are illustrative. On the Allied side, US forces experienced the most severe impacts of the contagion, with a rough 1 : 1 ratio of deaths from disease to battlefield injuries.30 French forces saw a much lower mortality ratio (roughly 1 : 6), and in British forces the ratio was approximately 1 : 10.31 Such variance in mortality contradicts Crosby’s postulate that all the belligerents were affected to the same degree by the pandemic. It also demonstrates that the British and French forces may have had more acquired immunity to the pathogen than their American counterparts.

Aside from the direct costs imposed by troop mortality, the effects of the influenza on military personnel included debilitation, decline of morale, and the diversion of the leadership’s focus from the prosecution of the war to containment of the contagion. Additionally, the death and debilitation of officers undermined force cohesion, planning, and execution, and reduced the capacity for effective reinforcement of divisions in the midst of battle. It also diminished the military’s capacity for the medical evacuation of ill personnel to hospitals, where health providers also succumbed to the virus. Thus, although the contagion did not utterly paralyze the machinery of war, it did slow military units down, notably diminishing their efficacy.

The United States

During the supposed first wave of pandemic influenza that appeared in the United States during the spring of 1918, initial mortality in military training facilities hovered around 10 per thousand. Such mortality eventually declined to mere 2.3 per thousand by mid September of 1918.32 However, the most virulent strain arrived soon after that point, and on October 11, 1918, mortality soared to 206 per thousand.33 The morbidity and mortality associated with the pandemic had profoundly deleterious consequences for the operational capacity of many US military units. In September 1918, US Brigadier General Charles Richard commented to his peers: “Epidemic influenza . . . has become a very serious menace and threatens not only to retard the military program, but to exact a heavy toll in human life. . . .”34 In the US experience, pandemic influenza induced absenteeism, loss of morale, and logistical chaos throughout the military infrastructure. The historian James Seidule argues that influenza had a significant negative effect on the morale of US forces and undermined their logistical cohesion during the conflict: “The flu sapped the strength and the morale of everyone in the AEF [and it] combined with malnutrition, inadequate clothing, and lack of sleep to create thousands of soldiers who suffered from combat exhaustion. . . . The result was an ineffective army with low morale.”35 Crosby describes the effects of the contagion on the US 57th Pioneer Infantry during their march to the naval vessels that would ferry them to France: “Some men stayed where they had sprawled. Others, almost as sick, struggled to their feet to keep up with their platoons, even throwing away equipment to avoid falling behind. No one was ever able to determine how much equipment or how many men the 57th lost on that march.”36

The rise of influenza had other pernicious effects on the war-fighting capacity of US forces, including logistical problems and the quarantine of US training camps. Eventually the virus resulted in suspension of the draft: “On September 26, with Pershing calling for reinforcements, with the AEF (American Expeditionary Force) pushing forward into the Argonne . . . the Provost Marshal General of the United States Army canceled an October draft call for 142,000 men. Practically all the camps to which they had been ordered were quarantined. The call up of 78,000 additional men in October had to be postponed. . . .”37 Crosby notes that the strains of the pandemic forced a 10 percent reduction in troop shipments to France. Furthermore, on October 11, 1918, “the War Department ordered a reduction in the intensity of training at all army camps. At the end of the month, the Chief of Staff in Washington wired General Pershing that the flu had stopped nearly all draft calls and practically all training in October.”38

The pandemic thereby undermined the American Expeditionary Force’s prosecution of the Meuse-Argonne Offensive, which began on September 26 and concluded on November 11, 1918. “When Pershing needed 90,000 replacement troops for his Meuse-Argonne campaign,” Byerly argues, [US Army Chief of Staff] March could provide him with only 45,000 because of the epidemic.”39 Indeed, during the Meuse-Argonne offensive, the estimated US morbidity from influenza was 68,760, as compared to 69,832 soldiers wounded by bullets and 18,864 who succumbed to gas.40 The comparative mortality statistics of the US 88th division are also illustrative regarding the effects of the pandemic. “The total of all combat losses for the 88th—killed, wounded, missing, and captured—was 90. The total of its flu cases during the fall wave was 6,845 [or] approximately one-third of the division. One thousand and forty five contracted pneumonia, and 444 died.”41

Alexander Stark, chief surgeon of the First Army, concluded that “influenza so clogged the medical services and the evacuation system, and rendered ‘ineffective’ so many men in the armies that it threatened to disrupt the war.”42 “By the War Department’s own account,” according to Byerly, “flu sickened 26 percent of the army—more than one million men—and accounted for 82 percent of total deaths from disease.”43

Given the malign synergy between the influenza virus and secondary sources of infection, one must include much of the subsequent mortality from infections (after October 1918) within the aggregate assessment of pandemic-induced mortality. Comprehensive assessments of mortality by the US War Department showed that the contagion actually killed more American troops than deaths from injuries sustained in combat. Specifically, while 50,280 American soldiers were killed in action, 57,460 died from pandemic influenza.44 The US War Department eventually estimated that the pandemic resulted in the loss of 8,743,102 person-days to influenza among enlisted personnel in 1918 alone.45 The US Navy was particularly affected by the destructive capacity of the pathogen: “All in all, the US Navy lost 4,136 of its officers and men to the flu and pneumonia in the last third of 1918. Despite the efforts of Germany’s undersea fleet, almost twice as many Americans soldiers died in the pandemic than as the result of enemy action in 1918.”46 US Naval statistics compiled during the course of the contagion indicate that circa 40 percent of naval personnel succumbed to the flu during 1918. Meanwhile, 361 per 1,000 US soldiers were admitted to hospital in 1918 for complications arising from influenza infection. “In total over 621,000 [US] soldiers caught the flu in 1918, upwards of one-sixth of the total number of American soldiers in World War I.”47 Furthermore, assessments of morbidity provide an indication of the aggregate impact of influenza on military effectiveness. The records of the Surgeon General of the US Army indicate that, of the 1 million men of the AEF who were hospitalized, circa 775,000 were hospitalized because of illness (influenza), while the remaining 225,000 were hospitalized for wounds incurred on the battlefield.48 Additional data indicate that 26 percent of the personnel in Army units were similarly debilitated by the pathogen.49

In the final analysis, US War Department records indicate that morbidity associated with the conflict saw 227,000 hospitalized for wounds incurred in battle, while over 340,000 were hospitalized for influenza.50 According to War Department records, the Army Surgeon General noted that debilitation and death from influenza had resulted in the loss of 9,055,659 days of manpower, with the result that almost two full divisions were out of action for the entire year 1918.51

Britain and France

According to Byerly, the influenza pandemic induced approximately 225,000 deaths among civilians in the United Kingdom.52 The British Expeditionary Forces saw approximately 313,000 cases (morbidity) of influenza during 1918, although incomplete records suggest that this estimate may be on the conservative side.53

Current estimates of French civilian flu-induced mortality are approximately 135,000, and France lost circa 30,000 soldiers to the virus over the course of the pandemic.54 Crosby argues that, in September 1918 alone, French forces in the combat zone exhibited over 25,000 cases of influenza, and that “the rear-area soldiers, not the men in actual combat, bore the brunt of the pandemic. The incidence of Flu in the French Army in the interior areas, for instance, was three to twelve times higher than in the French army at or near the battle-front.”55 This suggests that the rear camps may have acted as optimal vectors of viral dissemination, superior to transmission within the trenches.

Furthermore, the three waves of influenza that swept over the United Kingdom saw three distinct mortality peaks. The first peak occurred from July 6 to 14, 1918, the second from October 19 to November 23, 1918, and the third from February 8 to March 15, 1919.56 It is extremely significant to note that the waves of mortality that swept through Allied forces were not synchronous with those waves of mortality that swept across the citizens (and military forces) of the Central European states. The temporal variability of the mortality waves, hitting the Central Powers first, helps to further dispel the assertion that the influenza affected all the combatants in a roughly equivalent manner. As the German and Austrian data cited below indicate, the waves of mortality visited on those societies significantly preceded those that swept the United Kingdom, with possibly significant historical consequences.

Germany

The emergence of the virulent form of the contagion in the spring of 1918 coincided with the German Offensive of the Somme. During this period, the German Chief of Staff, General Erich Von Ludendorff, complained vociferously about the deleterious effects of the influenza on German military efficacy. The Somme offensive, which began on March 21, 1918, began to sputter in May as the virus increasingly debilitated German troops and crippled their units. In late June of that year, Ludendorff “noted that over 2000 men in each division were suffering from influenza, that the supply system was breaking down, and that troops were underfed. Infection spread rapidly. By late July, Ludendorff specifically blamed the pandemic for nullifying the German drive.”57 According to the historian Richard Bessel, “the influenza outbreaks in among the (German) troops in June and July 1918, left very great numbers of sick and wounded in their wake: of the 1.4 million German soldiers who participated in the offensives, over 300,000 became casualties between 21 March and 10 April, and the influenza epidemic in June and July alone affected more than half a million men; altogether between March and July 1918 about 1.75 million German soldiers fell ill at some point and roughly 750,000 were wounded.”58 Empirical data confirm Bessel’s assertions and Ludendorff’s protestations. Figure 3.2 illustrates the profound (and non-linear) increase in flu-induced mortality in the German Army. The pathogen-induced destruction was so profound, and the German physicians and nurses so overwhelmed by the dead and dying, that they were unable to keep track of the mortality as the first wave of the pandemic struck in the early days of June 1918, with over 15 percent of forces infected before the physicians were overwhelmed and ceased to register incidence.

Aside from the direct effects of mortality and morbidity, the flu also undermined the logistical efficacy of the German military: it adversely affected the effective function of the railways throughout the late summer and fall of 1918, impeding the timely distribution of materiel.59 “Influenza,” Crosby observes, “gummed up the German supply lines and made it harder to advance and harder to retreat. From the point of view of the generals, it had a worse effect on the fighting qualities of an army than death itself. The dead were dead. . . . But the flu took good men and made them into delirious staggering debits whose care required the diversion of healthy men from important tasks.”60

Ludendorff continued to lament the contagion’s pernicious effects on the German military effort. “Our army suffered,” he wrote. “Influenza was rampant. . . . It was a grievous business having to listen every morning to the chiefs of staffs’ recital of the number of influenza cases, and their complaints about the weakness of their troops if the English attacked again.”61 The death and debilitation arising from the pandemic also appear to have affected discipline and morale among German forces.62 According to Crosby, Ludendorff later “blamed the failure of his July offensive, which came so close to winning the war for Germany, on the poor morale and diminished strength of his armies, which he attributed in part to flu.”63

The public health records of the German military indicate that their forces were burdened by circa 1,543,612 cases of lethal influenza (then called “the Grippe”). Morbidity among German troops increased from a rate of 131,139 cases per annum in 1914 to 896,266 per annum in 1918,64 an increase of 683.18 percent over the course of the war.

According to the medical historian Fielding Garrison, infection rates ranged from 16 percent to 80 percent of soldiers, depending on the unit.65 By October 17, influenza had thoroughly debilitated the German forces and was raging all along the front. Bessel argues that the final lethal wave of influenza in October 1918 may have contributed to the ultimate collapse of the German war effort, and to the decline in effective governance by the state:

. . . in the last months of 1918 the military demobilization coincided with a sharp increase in mortality caused by the influenza epidemic: German civilian deaths shot up in October 1918 (when they were nearly two and a half times what they had been in September), remained extremely high in November, and returned to average wartime levels only in March 1919. Indeed, the months of October and November 1918 registered the highest civilian mortality rates in Germany for the entire war.66

With the notable exceptions of Bessel and Crosby, there has been a tendency among historians to discount Ludendorff’s accounts of the flu as given to wild exaggeration and exhibiting a tendency to deny accountability. However, Ludendorff’s accounts of the influenza are very much supported by empirical data culled from the Prussian State archives of the Staatsbibliothek in Berlin. (See figure 3.3.)

Waves of subsequent mortality due to secondary infection resulted from the destruction that the influenza had wreaked on its human hosts’ immune systems, particularly on the fragile tissues of the lungs. The most prevalent of these post-flu infections was tuberculosis, which claimed the lives of 40,043 German women in 1914 and 66,608 in 1918. Similarly, German female deaths resulting from pneumonia rose from 35,700 in 1914 to 74,468 in 1918.67 Collectively, this indicates that the conditions associated with the war resulted in increases of German female mortality of 66.34 percent for tuberculosis and 208.59 percent for pneumonia, both of which were typical post-flu secondary infections. Not only did the contagion kill productive members of German society and contribute to post-hoc infections and mortality; it also resulted in the profound and attenuated debilitation of survivors, to the extent that it undermined the productivity of the German workforce, and hence the German war effort, from the spring of 1918 to the collapse of the government in November of that year.68 Additional evidence for this drop in productivity comes from the dramatic declines in German coal production caused by influenza, as noted in the records of mines from that period.69 Such steep declines in the pivotal energy resource of the time would have had a profound negative effect on the capacity of the German military in particular, and on the resilience of the macro economy in general. Bessel has also noted that the rise of the influenza pandemic temporally coincides with the failure of the German military effort, the collapse of effective governance, and the advent of revolution.70 The data on morbidity and mortality support such assertions to a degree.

Austria-Hungary

The pandemic visited considerable suffering and destruction on the Central Powers and, as the data cited in figure 3.4 indicate, may have also resulted in limiting the martial power and the stable governance of Austria-Hungary.

The Austrian data (rather more complete than the German records) indicate that influenza struck Austria-Hungary in a single dramatic wave during late October and early November of 1918, whereas the rest of the world apparently suffered through three waves of influenza, beginning in the spring of 1918, each more lethal than the last. The Austrians apparently only suffered the visitation of the last and deadliest viral wave, which cut through the population in late October 1918.

The data suggest that the Austrians witnessed a previously unacknowledged regional epidemic of considerable influenza-induced mortality during the first and second quarters of 1917, significantly predating the viral waves that began in the spring of 1918 in the United States. This revelation suggests that a precursor epidemic, of unknown origins, apparently swept through Austria (and perhaps other regions of Central Europe) in the spring of 1917. This provides empirical evidence to reinforce John Oxford’s hypothesis that the virus may have emerged in sporadic epidemic form before 1918.71 What is apparent in the Viennese case is that the Austrian people did not experience the two increasingly lethal waves of pandemic influenza that swept the world in the spring and summer of 1918 but suffered a lingering pandemic. From an epidemiological viewpoint, the lack of exposure to these two waves in early 1918 may have inhibited the Austrian population’s development of any significant immunity to the final genetic variant of the pathogen, thereby increasing their vulnerability to the third wave.

Such empirical evidence of a Austrian precursor epidemic that generated significant mortality in the spring of 1917 is of profound epidemiological significance. It explains how the Austrians may have developed partial immunity to the viral waves of spring and summer 1918, as a direct result of that prior exposure. However, they apparently remained immunologically naive to the genetically novel and exceptionally lethal variant of the virus that appeared in the fall of 1918 and generated such enormous mortality in the fourth quarter of that year.

Further, one must recall the pathogenic connectivity between influenza, which destroyed the hosts lungs and immune system, and the opportunistic infections (such as viral and bacterial pneumonia) that then debilitated and often killed the weakened host. Additional data from Austria indicate that the influenza pandemic generated a subsequent wave of pneumonic mortality throughout the Austrian people. This secondary wave of infection significantly augmented the mortality generated by the final viral wave of influenza. Furthermore, it is interesting to note that an earlier peak in pneumonia-induced mortality appears to be synchronous with the observed “first wave” of influenza mortality that struck Austria in the spring of 1917.72 (See figure 3.5.)

The historical record indicates that the final viral wave of influenza (and the nigh synchronous wave of deaths from pneumonia) immediately preceded the fragmentation of the Empire. Specifically, within two weeks of the final wave of the virus’ striking Vienna, the Austro-Hungarian Empire underwent utter political disintegration. Thus, mortality from the “third wave” constituted an exogenous shock of some magnitude, and one that may have served as the proverbial “straw that broke the camel’s back.” It is quite certain that influenza was not the sole agent responsible for the dissolution of the empire. However, the influenza pandemic undoubtedly functioned as a powerful stressor to shatter the rotten and tottering foundations of the institutions of the empire, which had been successively eroded by years of war.

Pandemic Influenza’s Impact on World War I

With the notable exceptions of Crosby, Byerly, and Bessel, medical historians have perpetuated the ignorance of the impact of influenza on the war effort, and certainly on the outcome of the war. The historians Byron Farwell, Jennifer Keene, and Robert Ziegler have argued that the pandemic compromised the effectiveness of military forces during the war to a certain extent, but they have not gone so far as to argue that it had any effect on the outcome of the conflict.73

Given that the waves of pandemic influenza (as determined by pathogen-induced mortality) struck Austrian and German society (and their military forces) before they struck British society, we might expect that influenza debilitated the war effort of the Central Powers more than that of the Allies. Thus, it is variation in mortality and (perhaps even more important) in timing that indicates that Crosby was premature in concluding that all combatants were affected in the same manner. The balance of evidence accumulated herein suggests that the pandemic affected the combatant’ militaries, governments, and societies in rather different ways, and that it may have contributed to the defeat of the Central Powers.

As the data indicate, the contagion eroded the capacity and efficacy of affected militaries, diminishing their optimal functionality. The pandemic’s infliction of such debilitation and death on all the protagonists in the war may have effectively brought the conflict to an early end, as military institutions became increasingly sclerotic and ineffective. Moreover, there is some evidence (borne out in the German mortality tables) that the pandemic greatly impeded the German offensive of the spring of 1918, which, if successful, would have resulted in victory for the Central Powers. The second and third waves also diminished German martial power throughout the summer and fall of 1918.

Thus, the epidemic may have prevented a German victory, extended the war, and ultimately assisted in forcing the Central Powers to the table to negotiate peace at Versailles. On October 6 and 7, 1918, at the height of the influenza pandemic in Central Europe, the governments of Germany and Austria sent notices to US president Wilson requesting an armistice and peace negotiations based on Wilson’s proposed fourteen points. However, perhaps the most important (and previously unexplored) point is that the visitation of the final lethal viral wave on the immunologically naive population of Austria resulted in widespread death and debilitation and in sclerosis of governance, and ultimately contributed to the collapse of the empire. By articulating this hypothesis I hope to stimulate some debate as to the role of the influenza in the collapse of Austria-Hungary and in the demise of Imperial Germany.

Scapegoating

The history of disease, particularly during its manifestation in European societies, is riddled with fear-induced desires to target minorities as the carriers, instigators, or vectors of disease transmission. It seems a common frailty that humans find it expedient to blame the psychological “other” for visitations of contagion, even pandemic influenza. Despite the fact that the strain may have originated in Kansas (or Austria, for that matter), it was subsequently labeled the Spanish Flu. This designation resulted from the fact that Spain, a neutral party during the conflict, was not actively engaged in the censorship of the reporting of the epidemic at that time. Citizens on the Allied side began to envisage the contagion as some nefarious and demonic weapon conjured by the Germans to poison the people of the Allied countries. For example, one excessively passionate argument went as follows: “Let the curse be called the German plague. . . . Let every child learn to associate what is accursed with the word German not in the spirit of hate but in the spirit of contempt born of the hateful truth which Germany has proved herself to be.”74 While this effect apparently did not result in attacks on demonized minorities (beyond the obvious organized violence directed toward the Central Powers), the central function of contagion-induced stigmatization appears to continue to hold in this case.

Governance

At the domestic level, influenza had a sclerotic effect on governance within severely affected countries, overwhelming the capacity of the state (and often the society) to deal with the debilitation and mortality generated by the contagion. Crosby estimates that the influenza pandemic of 1918-19 generated at least 550,000 excess deaths in the United States (i.e., over and above those deaths that would typically result on an annual basis).75 As one would expect, one of the first sectors of US society to be overcome was the health sector. Hospitals did not possess the requisite surge capacity to deal with such a huge influx of ill patients. Specifically, hospitals did not possess the necessary beds and supplies, nor did they have adequate reserves of medical personnel (nurses and doctors) on hand to deal with the surge in infected civilians. Moreover, lacking adequate protection, many health providers themselves succumbed to the illness, and thereby became an additional burden on those who remained healthy. The contagion also undercut the timely and effective delivery of other public goods by the state to the people, including essential services such as communications. Crosby notes that “eight hundred and fifty employees of Bell Telephone Company of Pennsylvania stayed home from work on October 7, 1918” and that “on the next day Doctor Krusen of the Department of Health and Charities authorized the company to deny services to any persons making unessential calls, and it presently did so in a thousand cases.”76 This sclerotic effect of the contagion also impacted the police and family services. The pandemic resulted in rampant absenteeism among police officers, firemen, garbage collectors, and social workers to care for children who had lost their parents.77 Further complicating the situation was the inability to bury the dead promptly and effectively:

The essential service which came closest to collapse in Philadelphia was (morticians). Unless morticians are able to fulfill their duty, two things happen. One, bodies accumulate, creating the possibility of secondary epidemics caused by the various organisms that batten on dead flesh. Two, and more immediately significant, the accumulation of corpses will, more than anything else, sap and even break the morale of a population. When that happens, superstitious horror thrusts common decency aside, all public services collapse, friends and even family members turn away from one another, and the death rate bounds upward.78

The discord associated with the pandemic was certainly not relegated to the environs of Philadelphia, but distributed throughout the United States. In the city of San Francisco there were acute shortages of medical personnel, police, communications personnel, educators, and even garbage collectors. “The Sanitary reduction works shut down when only 11 of its normal staff of 56 showed up for work.”79

While the state’s ability to respond was significantly curtailed by the contagion, successful adaptive response came in the form of the mobilization of civil society. It was this galvanic response of non-state organizations, religious, social, economic, and political, that enabled American society to overcome the ravages of the pandemic.80 Regarding this response by society, it seems reasonable to argue that the civil cohesion generated through the prolonged war effort (notably the Civil Defense Associations) empowered civil society to the extent that it was capable of dealing with the widespread death and disruption generated by the pathogen. It should be noted that different communities responded with varying degrees of effectiveness. San Francisco, for example, was plagued by particularly inept responses, by both civil society and the state. “Despite preliminary planning, organization never caught up with the flu until it had passed its peak. No local Council of National Defense arose to coordinate anti-pandemic forces; no central clearing house to process all calls for assistance . . . was ever created, and San Franciscans ran their doctors ragged checking on cases that needed no professional attention.”81

The pandemic serves as a prime example of emergent properties, and the Oxford hypothesis supports this line of reasoning. The lethal pandemic influenza of 1918 likely derived its intensity from a combination of the conflict’s constituent attributes (and their side effects). Such pernicious factors included the dense troop populations that moved rapidly and continuously around the world (functioning as highly efficient vectors of transmission), coupled with poor nutrition that undermined immune systems, the highly unsanitary conditions of the trenches and military camps, and a novel zoonosis (H1N1 avian influenza). Those permissive conditions, which resulted in rapid viral transmission from host to host, facilitated the evolution of traits of lethality in the virus, resulting in a highly contagious and lethal influenza pandemic. Individually, each one of these constitutive variables may have not generated any significant effect, but when combined in this fashion, led to one of the greatest global public health disasters in recorded history.

Ultimately, the balance of evidence from Germany, Austria, and the United States suggests that the 1918 influenza had various effects on state capacity in affected polities. One obvious effect was that the morbidity and mortality generated by the influenza pandemic generated profound institutional sclerosis. The strongest evidence for this emanates from the extensive problems that became manifest in both the Allied and German military forces during 1918. Other bureaucracies exhibited sclerosis in the United States, particularly those that provided public services such as health care, communication, law enforcement, sanitation, and so forth. Although the preliminary evidence presented here suggests that pandemic influenza did significantly impede the optimal function of state institutions in 1918, further cross-national historical research is required to validate this hypothesis.

The Swine Flu Affair and Its Repercussions

In the wake of the great pandemic of 1918-19, the twentieth century witnessed additional (and relatively minor) pandemics during 195782 and 1968.83 The processes of emergence of pandemic influenza are cyclical and thus, contain a degree of periodicity.84 However, the Swine flu affair of 1976, which emanated from initial cases of flu-induced mortality in Fort Dix, New Jersey, generated problems in US domestic response to contagion that persist to this day. Despite dire warnings regarding the emergence of a novel strain of highly pathogenic influenza, the 1976 flu failed to generate the high levels of morbidity and mortality that had been predicted. However, profoundly negative repercussions did result from the vaccination program that was authorized by the US government, which rushed though a prototype vaccine without adequate testing before mass dissemination.85 Owing to the faulty production and insufficient testing of that vaccine, the provision of such vaccinations to the American people resulted in a significant number of people becoming afflicted with Guillain-Barre paralysis.86 The result was a storm of litigation against the manufacturers, which ultimately resulted in the courts’ awarding huge damage settlements to plaintiffs. Ultimately, owing to the litigation, the major US-based developers of vaccines were forced to relocate to Canada and to Europe. This situation persists in the twentyfirst century, greatly complicating domestic US ability to respond to a future influenza pandemic. As a result, the US is now almost completely dependent on vaccines produced abroad, and on the honoring of contractual obligations in the face of a global health emergency.

The Future

There is great concern about the geographical spread and the persistence of the H5N1 Avian Influenza that appears to be endogenized within the human ecology of Southeast Asia.87 The pathogen appears to be a highly lethal88 zoonosis with the unusual property of being able to jump directly from its natural avian reservoirs into human hosts.89 The virologists Jeffrey Taubenberger and David Morens argue that, despite our armamentarium of vaccines and antivirals, an emergent influenza pathogen that exhibited lethality of the same magnitude as the 1918 virus would kill more 100 million people today.90

In the domain of international commerce, the current strain of avian influenza has already inhibited flows of goods and people to a minor extent. During 2006 the European Union banned imports of poultry and bird products from Bulgaria, including wild birds, eggs, farmed and wild feathered game, and hatching eggs. Further, a regional ban was applied to all imports of poultry meat, eggs, and products from wild fowl. Current EU policies do not provide for compensation to farmers who incur losses as a result of declining public confidence in the safety of poultry.91 In the European context, the arrival of the virus has already resulted in reduced consumption of poultry, generating hundreds of millions of dollars in losses for that industry.92 According to the World Health Organization, the disease has already cost 300 million farmers more than $10 billion as a result of its spread through poultry.93

Moreover, the next pandemic has implications for the food security of affected countries, as it has already resulted in the culling of millions of birds. Joseph Domenech, chief veterinary officer of the United Nations Food and Agriculture Organization, has cautioned that the spread of the epidemic may undercut nutrition: “If a poultry epidemic should develop beyond the boundaries of Nigeria the effects would be disastrous for the livelihoods and food security of millions of people.”94 This may explain why Nigerian officials were aware of the pathogen within their avian populations for 19 days before informing the public and the international community.95 Obviously, such deliberate obfuscation and delay can only undercut international attempts at containment.

In 2006, European countries reported the arrival of the H5N1 virus, which has made sporadic appearances in the United Kingdom, France, Germany, Austria, Greece, Italy, Bulgaria, Poland, and Slovenia. Further, the disease is now apparently established in Russia, Ukraine, Romania, Turkey, Bulgaria, Croatia, Egypt, and Azerbaijan (all non-EU countries). EU governments continue to discuss plans to create a pan-European program to vaccinate poultry. In mid February 2006, EU governments announced a program of strict precautionary measures for containment. Specifically, they imposed a general rule that applies and enforces a quarantine and surveillance zone of about 10 kilometers around areas where the virus has been detected.96

Containment of the pathogen is likely to be problematic throughout much of Africa, since the slaughter of poultry stocks by government forces is typically not accompanied by reimbursement from the state in this region. As a result, farmers—particularly in Nigeria—have a significant economic disincentive to report unusual avian mortality. This limits surveillance and the execution of containment strategies. Further, many governments in Africa exhibit exceptionally low levels of fiscal capacity, and therefore may be unable to make effective compensation payments. According to former WHO Director Jong Wook Lee, the international community should create a mechanism to cover excess costs associated with such compensation, in order to ensure accurate surveillance and compliance throughout less developed countries.97

This situation is complicated by those societies that harbor a legacy of mistrust between civil society and the government, particularly those polities that are in the process of transition from authoritarian rule to nascent democracy. Moreover, the lack of government legitimacy, and education of the population, are also hampering efforts to control the influenza in Nigeria. According to the journalist Dulue Mbachu, “a wall of distrust between the government and most of the population is proving a major obstacle to fighting bird flu in Nigeria. The campaign is also hampered by poor infrastructure, lack of resources, and vast distances. In Nigeria, after decades of misrule by corrupt military and civilian regimes, the 70 percent of the population with little education or income has grown wary of all officialdom.”98

In the United States, Health and Human Services Secretary Michael Leavitt announced in 2006 that he had authorized the National Institutes of Health and the Centers for Disease Control and Prevention to prepare a second vaccine to counter the H5N1 virus, based on the fact that the prior vaccine was based on samples taken from Thailand in 2004. Health officials now believe that the virus has undergone significant mutation since 2004, and that the form now circulating in Africa and Europe may exhibit significantly greater genetic variance than the prior variant.99 Although the United States currently possesses a stockpile of 5 million doses of Tamiflu (oseltamivir), in March 2006 it ordered 12.4 million more doses.100 Unfortunately, recent evidence suggests that certain strains of the virus are highly resistant to oseltamivir, and so the significant expense incurred in stockpiling the compound may not in fact result in the expected positive dividends of protecting the lives of the US public.

Early in 2006, the US Congress authorized $3.8 billion for the purchase of more vaccine and Tamiflu from the Swiss firm Roche and from the British firm GlaxoSmithKline.101 The central problem emanates from the global competition to procure a rather limited supply of anti-viral prophylaxis, while global production capacity remains inadequate. Compounding the problem, the US federal government has asked the states to create their own individual stockpiles, which may encourage hoarding by wealthier states (such as California and New York). There is currently no federal legal architecture that can compel these states to share their supplies in a cooperative manner to maximize efficiency should a pandemic occur. And there are no substantive international mechanisms to ensure the cooperation of sovereign states.

However, the necessity of developing protocols for international cooperation to combat emerging H5N1 strains has percolated into the upper echelons of the policy-making community, and Paula Dobriansky, former US Undersecretary of State for Global Affairs, has wisely argued for greatly increased international cooperation on the issue.102 Domestically, the current US strategy is to rely on actions taken abroad to contain the proliferation of the virus, but that is a dubious strategy on several grounds. First, there is enormous variance in the endogenous capacity of foreign states to conduct effective pathogenic surveillance and containment, ranging from the relatively robust capacity of the G-8 countries to the almost non-existent public health infrastructural capacity of states such as Haiti, Ghana, and Cambodia. The situation is exacerbated by perpetually feeble international regimes (including the revised International Health Regulations) and poorly funded international institutions (the WHO). While the WHO was reasonably effective in assisting sovereign states to contain SARS in 2003 (see chapter 5), in recent years the organization has witnessed the re-emergence of poliomyelitis virus in Africa and its spread back to South Asia,103 an inability to contain the burgeoning HIV/AIDS pandemic, and the continued proliferation of malaria, tuberculosis, and hepatitis around the world.

In the context of such weak international institutions, and with states serving their own material self-interests, we are likely to see less than optimal international cooperation in the face of a highly pathogenic pandemic influenza. The agents of international organizations have admitted as much. Mike Ryan of the WHO recently warned that global capacity for containment of the emerging pandemic is insufficient: “We truly feel that this present threat and any other threat like it is likely to stretch our global systems to the point of collapse.”104 Joseph Domenech, head of the UNFAO’s Animal Health Service, complained that the developed world had not done enough to contain the spread of the pathogen throughout the developing world, where countries have insufficient capacity for surveillance and control: “In 2004 we said there will be an international crisis if we don’t stop it in Asia, and this is exactly what is happening two years later. We were asking for emergency funds and they never came. We are constantly late.”105

The primary concern, then, is that a pandemic exhibiting pathogenicity similar to that of the 1918 virus would overwhelm institutions of governance in the G-8 countries, and to an even greater extent in the less developed countries. Arguably, the globalization of tightly coupled economic systems has made us more vulnerable to pathogen-induced disruptions than were our forebears in 1918. Furthermore, such vulnerability is exacerbated by the greatly increased speed of pathogenic transmission, courtesy of modern transportation technologies. Moreover, the SARS epidemic illustrated that the modern media and telecommunications technology may exacerbate economic damage through its rapid diffusion of anxiety, fear, and panic.

Within the United States, the capacity of institutions (both at the state and federal levels) to mitigate the negative externalities associated with an influenza pandemic is very much in doubt. Let us briefly examine the shortcomings at the level of the individual states. According to the dictates of the US Constitution, the individual states possess the legal authority to deal with crises in the domain of public health. However, there is enormous variation among the states in endogenous capacity, including human capital resources (i.e., trained personnel), pathogenic surveillance capacity, fiscal resources, health infrastructure, surge capacity in hospitals, and administrative preparedness for health emergencies. One might simply compare the state of New York, with its vast post-9/11 resources and augmented preparedness, to poorer states, such as North Dakota and New Mexico, which struggle to find the fiscal resources to conduct basic public health surveillance.

At the federal level, the United States’ capacity for response is inhibited by a number of factors. First and foremost is the lack of federal powers to deal with a public health emergency, which became quite evident during the Andrew Speaker affair during May and June of 2007. In that particular case, the CDC found itself beholden to the State of Georgia in its attempts to limit the movement of an individual infected with a rare and exceptionally drug-resistant strain of tuberculosis. Certainly, the US Congress could claim jurisdiction over public health emergencies through exercise of the Commerce Clause in order to grant federal bureaucracies such powers.106 A further problem results from the chronic fragmentation of oversight of health issues among (and within) the various federal bureaucracies, which possess degrees of overlapping jurisdiction in the domains of surveillance, management, and pathogenic containment. For example, the CDC often competes with the NIH within the Department of Health and Human Services over attribution, access to data, and funding. In the face of an avian influenza pandemic, HHS would have to cooperate not only with the states but also with the Departments of Homeland Security, Defense, Agriculture, Interior, Transportation, Commerce, and State. At present there is no cabinetlevel official tasked with coordinating a national response, and, as the exceptionally inept federal and state governmental responses to Hurricane Katrina indicated, cooperation between federal bureaucracies and between the federal government and the states cannot be taken for granted. Thus, to optimize the domestic response, a pandemic flu coordinator should be designated at the cabinet level.

In view of the problems evident in US domestic disaster response, the role of civil society remains integral to the provision of effective response in the face of pandemic influenza. In the face of the 1918 contagion, local Civil Defense Associations (i.e., trained civilian volunteers) provided information and assisted beleaguered health-care providers. As Putnam has documented, the gradual erosion of civil society and the consequent erosion of social capital in the United States107 bode ill for our collective capacity to respond. However, recognizing that civil society may constitute a powerful force for positive intervention, the country should invest in the promotion of preparedness through local civic organizations that can assist the government during such a crisis.

The case of the 1918 pandemic influenza is consonant with a republican reformulation of Realist theory. The pandemic represented a direct threat to the material interests of all countries, political suppression of data often prevented other states from knowing the conditions of affected states, rational decision-making was largely absent, and international cooperation on the issue was non-existent, despite the complex biological interdependence of affected polities.