The Lasting Effects of the 1918 Influenza Pandemic

[I’ve never put a trigger warning on a post before but given the current situation the information here is potential upsetting to anyone expecting a child. I do not think that the current pandemic will be as bad as the 1918. I am also hopeful that the weather will work in our favor and that, as Tyler argued, America will start to work. Do also read my post, What Worked in 1918-1919 for a more positive message.]

The 1918 influenza pandemic struck the United States with most ferocity in October of 1918 and then over the next four months killed more people than all the US combat deaths of the 20th century. The sudden nature of the pandemic meant that children born just months apart experienced very different conditions in utero. In particular, children born in 1919 were much more exposed to influenza in utero than children born in 1918 or 1920. The sudden differential to the 1918 flu lets Douglas Almond test for long-term effects in Is the 1918 Influenza Pandemic Over?

Almond finds large effects many decades after exposure.

Fetal health is found to affect nearly every socioeconomic outcome recorded in the 1960, 1970, and 1980 Censuses. Men and women show large and discontinuous reductions in educational attainment if they had been in utero during the pandemic. The children of infected mothers were up to 15 percent less likely to graduate from high school. Wages of men were 5–9 percent lower because of infection. Socioeconomic status…was substantially reduced, and the likelihood of being poor rose as much as 15 percent compared with other cohorts. Public entitlement spending was also increased.

At right, for example, are male disability rates in 1980, i.e. for males around the age of 60, by year and quarter of birth. Cohorts born between January and September of 1919 “were in utero at the height of the pandemic and are estimated to have 20 percent higher disability rates at age 61…”.

Figure 3 at right shows average years of schooling in 1960; once again the decline is clear for those born in 1918 and note that not all pregnant women contracted influenza so the actual effects of influenza exposure are larger, about a 5 month decline in education, mostly coming through lower graduate rates.

Higher disability and lower education translate into greater government payments as show in the final figure below. Almond labels these welfare payments which might be slightly misleading–these are Social Security Disability payments in 1970. Here’s Almond:

Average payments to women and nonwhites in 1970 are plotted in figure 8. The average welfare payment was 12 percent higher for both women and nonwhites born in 1919, or approximately one-third higher for children of mothers who contracted influenza. When we focus on quarter of birth, it is apparent that these increased payments are generated by high payments to those born between April and June of 1919.

Note that men and women who were especially disabled could have died before 1970 and so these are lower bounds on the disability impact.

Fetal exposure seems to be the key as Almond tests for and rejects other possibilities. The 1918 kids, for example, seem about the same as the 1920 kids so it’s not that the flu killed off the weak kids in 1918.

Almond was interested in the 1918 pandemic not simply as a historical episode but to make the case that infant health and infant health programs have high benefit to cost ratios, a still relevant lesson.

Hat tip: Wojtek Kopczuk.


Nice article on the very long term effects of children. At this point we should also still be concerned about possible residual effects of those who have had Covid-19. With the flu once you recover you generally go back to your baseline level of health. I dont think we know yet if that is true for Covid. If there are long term pulmonary residual effects then the costs of this outbreak will be higher than we are already assuming.


Given that Alex does not cite similar IQ effects for any flu outbreak more recent than a hundred years ago, it seems likely that the Spanish Flu was an outlier in that regard.

Hi, this article about The Lasting Effects of the 1918 Influenza Pandemic is very useful
and inspirational. That is why I want to help me to leave you a way to get rid of herpes:
Great success with this site!

For most younger people, this virus stays in the nose and throat and behaves like a cold. I would expect no impact to a fetus. In the cases where it proceeds to the lower respiratory tract, it's possible but I don't think we really know yet.

How about giving information from the CDC website rather than speculating based on 1918:

"Q: Are infants born to mothers with COVID-19 during pregnancy at increased risk for adverse outcomes?

A: Based on limited case reports, adverse infant outcomes (e.g., preterm birth) have been reported among infants born to mothers positive for COVID-19 during pregnancy. However, it is not clear that these outcomes were related to maternal infection, and at this time the risk of adverse infant outcomes is not known. Given the limited data available related to COVID-19 during pregnancy, knowledge of adverse outcomes from other respiratory viral infections may provide some information. For example, other respiratory viral infections during pregnancy, such as influenza, have been associated with adverse neonatal outcomes, including low birth weight and preterm birth. Additionally, having a cold or influenza with high fever early in pregnancy may increase the risk of certain birth defects. Infants have been born preterm and/or small for gestational age to mothers with other coronavirus infections, SARS-CoV and MERS-CoV, during pregnancy.

Q: Is there a risk that COVID-19 in a pregnant woman or neonate could have long-term effects on infant health and development that may require clinical support beyond infancy?

A: At this time, there is no information on long-term health effects on infants either with COVID-19, or those exposed to the virus that causes COVID-19 in utero. In general, prematurity and low birth weight are associated with adverse long-term health effects."

Here is the link:

Nothing qualifies economists to be neonatal specialists.

This is one of those taboo subjects that neonatal specialists tend tho avoid. Google Scholar -- or the March of Dimes "39 weeks" videos and pages -- are blunter. Being born prematurely for any reason, including doctor-advised induced birth for preeclampsia risk, has substantial negative effects on the baby's brain development, lung development, vision, and other factors. Even 37 weeks is too early, despite its being branded as "full term" in the U.S. and certain European countries. For instance, you lose 2 or 3 IQ points per premature week short of 39 or 40 weeks.

Understandably, nobody wants to convey this information to a pregnant woman.

I've pointed this out before in the context of the "reverse flynn effect". Doctors like to induce at 39 or 40 weeks if a woman is over 35 to reduce the (already low) risk of still birth, artificially shortening gestation. There may be a subtle effect on brain development even at that date.

"For instance, you lose 2 or 3 IQ points per premature week short of 39 or 40 weeks."

Is that linear? I've seen a chart which showed a substantial difference for 4+ weeks premature, but negligible differences for 4 weeks or less. Which presumably is one of the factors on why Doctors tend to induce early.

I'm not sure about the linearity, but a baby's brain grows 50 percent bigger in the last four weeks, 35 to 39/40 weeks. See the graphic below the fold in this Wall Street Journal article:

Bill: "How about giving information from the CDC website rather than speculating based on 1918.."

CDC website to which Bill directs us: "At this time, there is no information..."

Yes, that is correct, so please read Alex's trigger comments: "the information here is potential upsetting to anyone expecting a child. I do not think that the current pandemic will be as bad as the 1918."

Apparently you do not read, nor do you comprehend. I would suggest that Alex would have said: CDC says....

This coronavirus doesn’t seem to affect young people as much in contrast with the Spanish Flu, which disproportionately affected young people. Hopefully that means impacts on infants and unborn children will also be limited.

The long-term impact that I’m more concerned about is that the travel restrictions never fully go away, just like the post-9/11 security restrictions never fully went away. That could depress the world economy permanently (and perhaps explains the fall in long-term interest rates with the 10-year bond now at 0.5% even though this disease itself will almost certainly be gone in 10 years).

Bonds are at record lows because every asset class has systemic risk and institutions need a liquid place to park huge volumes of money to preserve their principle until they can identify where to put it for growth.

As far as I can see, Almond has overlooked whether being the offspring of a now-dead father (who died between the child's conception and birth) has an effect separate to in utero exposure, which seems a correctable oversight. My great-grandfather was taken from a very young family by the Spanish flu, leaving them in considerable hardship. I imagine it would have been even worse if my great-grandmother had been expecting.

Good point. One might think that the 1918 babies had some dads who went off to war and died, which could show up in the 1918 v 1920 comparison. But what proportion of missing dads in each year?

Related question: The first few months *after* a 1919 birth had to be pretty unsettled as well. Missing relatives, less support for mom - would Almond have been able to sort that out, and how likely is the impact to be dramatic?

Wikipedia - In the United States, the disease was first observed in Haskell County, Kansas, in January 1918, prompting local doctor Loring Miner to warn the U.S. Public Health Service's academic journal. On 4 March 1918, company cook Albert Gitchell, from Haskell County, reported sick at Fort Riley, an American military facility that at the time was training American troops during World War I, making him the first recorded victim of the flu.[41][42][43] Within days, 522 men at the camp had reported sick.[44] By 11 March 1918, the virus had reached Queens, New York.[40] Failure to take preventive measures in March/April was later criticised.[45]

In August 1918, a more virulent strain appeared simultaneously in Brest, France; in Freetown, Sierra Leone; and in the U.S. in Boston, Massachusetts.

And since those dates are not really accurate, at least regionally, such research could possibly be refined to see if the results hold up for Boston babies born in September 1918

There is also a link between the flu and atherosclerosis. Some have claimed that the cardiovascular disease epidemic that peaked in the 60's was the echo of the 1918 Spanish flu pandemic. Coronavirus is probably different since the symptoms in young people are relatively mild.

Hi, I am Gary Wilcox, a fifth-generation Republican voter. No one in my family has voted for a Democrat since the 1830s. I have always voted straight Republican tickets. This year, however, I will do so nothing I never thought I would: vote for a Democrat. My family and friends are astonished. Well, I decided I can not vote for Trump again after he mismanaged America's response to the coronavirus crisis and caused tens of deaths with (unfortunately) more to come. Could I bring myself to vote for a Democrat though? That remained to be seen. I had to study the options. There is so much at stake! And, man, let me tell you that, Joe is different. He is not another of those crazy elitists Democrats try to impose on us (remember Hellary? LOL). He is a down-to-earth, affable, tell-it-like-it-is American like you and me. He is an experience, reliable, honest public official. For all those reasons and more, I have made my decision: I will vote vote for Joe. We know Joe!

Thanks for the PSA, Thiago.

Oh, no. I am Mr. Wilcox, a common American who is sick and tired with business as usual politics.

>I do not think that the current pandemic will be as bad as the 1918.

Really, now?

The standard flu kills 30,000 to 70,000 Americans EVERY YEAR.

Coronavirus stands at 22 people.

Such a bold take you have there. Thanks for the trigger warning!!

And less than three weeks ago, it stood at zero.

We don't even have enough tests so both of you are talking out of your asses.

The first two American deaths from covid19 were on Feb. 26, 2020, which is less than 3 weeks ago. However, they were not recognized as being caused by covid19 until March 3, with the first (erroneously) reported covid19 death in the U.S. being reported on Feb. 29.

There is no question that lack of adequate testing resources in the U.S. makes accurate data analysis difficult. It is possible that with further examination, the date and number of fatal covid19 victims will change, in a historical sense.

According to the HHS guy we have tested tens of thousands, hundreds of thousands, millions!!!

The closing paragraph of the STAT article Some simple math offers alarming answers:

Unwarranted panic does no one any good, but neither does ill-informed complacency. It’s inappropriate to assuage the public with misleading comparisons to the seasonal flu or by assuring people that there’s “only” a 2% fatality rate. The fraction of cases that are severe really sets Covid-19 apart from more familiar respiratory illnesses, compounded by the fact that it’s whipping through a population without natural immune protection at lightning speed.

The good news is that a baby born in 2020 will be 10 years old in 2030, 20 years old in 2040... (you can see where I'm going with this) and 30 years old in 2050, which will be radically different medically from 1928, 1938 and 1948.

The Spanish flu was unusual in that young people suffered the most.

Vaccination and medicine wasn't quite there in those days. Even today's generation has the benefit of vaccinations, antibiotics, and being descended from pandemic survivors.

You may want to look up the meaning of cytokine storm, and its effects.

How's it stack up to a Sharknado?

In this corner, the Spanish Flu cytokine storm, having killed millions and millions.

And in this corner, a silly joke that isn't killing anyone reading it.

It was very specific to that outbreak. We keep comparing Coronavirus to the Spanish flu, but the 1918 pandemic was atypical,

Each new pandemic is atypical, virtually by definition.

Comparisons may have greater or lesser power, but the Spanish flu was the last recognized global pandemic, and was spread in at least somewhat similar fashion to covid19. Other similarities may or may not prove to be worth comparing at this point.

They all vary. They are not all atypical. And there have been two flu pandemics since 1918. They just weren't as large.

I would be interested in hearing Alex/TC's opinion on what the tipping point should be for US to start implementing emergency protocols (e.g., forced cancellation of mass social events, limiting citizen travel).

Not too long ago TC had a post about "responder types" (react now vs. watch and wait). Most of the time watch and wait is most prudent, but occasionally the "react now's" get it right. I'm typically a watch and wait person, but I am wondering if we need to hit panic button. I'm a physician at one of the largest academic, tertiary referral hospitals in the country (with one of the the largest geographic catchment areas). We have NOT had a Covid case yet. We are ALREADY full, all of the time. Our ICU's are always full (we have one of the highest # of ICU beds in the country). We could not handle even a 25% increase in the number of ICU admissions (which are typically transfers from outlying community hospitals). This will spill over to those hospitals who have much smaller ICUs and much fewer ventilators and respiratory therapists.

I think at this point, most people agree that Covid won't be contained. But extreme distancing measures could slow the spread, which would slow the number of new cases, and be less likely to overwhelm hospital systems. The question is when do you implement these measures, if you wait too long, less likely to be effective. The more upstream, however, bigger and longer the impact on economy.

Are you sure this is true, particularly about such things as ICU resources? There is another doctor here who is constantly talking about how America has so much more extra capacity compared to any country with a single payer system.

Yes. It's not even really about ICU "Beds" b/c floor beds can be converted. It's the ventilator machines, staffing, and other ICU-related equipment. Our hospital simply does not have enough machines to accommodate >25% cases requiring advanced respiratory support.

The bigger issue again is smaller, community hospitals. Once tertiary referral centers can't take any more patients (which could happen quickly), community hospital will have to provide this care. They definitely don't have a large # of excess supplies, not too mention the staffing.

I am not sure what "extra capacity" America has. I guess we have the ability to spend lots of $, but at this point cost really won't be the issue. Is supplies and staffing that will be overwhelmed. And as many have already pointed out, global supply chains won't keep up.

I would be interested in how we can increase capacity to handle health crises by taking existing human and other capital and cross training it to handle health emergencies.

For example, should we be training National Guard soldiers on how to transport infectious persons or how to set up outdoor tents near hospitals to screen individuals.

As part of military training, should recruits be trained on how to handle infectious persons, how to prevent spread, how to wear hazmat suits.

Should we close the military bases that, for some reason (local politics) never get closed, and convert them to places to handle surge medical events or other natural disasters? As a palliative to the adjoining community, build out their hospital infrastructure.

Should we be putting people in cages at the border when there is a risk of spread among the persons in the cage which then spreads to boarder enforcement personnel.

Will parole be better than imprisonment if prisons could become hotspots.

Is your hospital doing thing to increase capacity?

Our hospital network has already set up plans for what we need to do if we need to increase ICU capacity. However, Justin is correct that in most tertiary care hospitals the ICUs are full almost all of the time. Plus, it isn't just bed capacity. You will have critical staffing shortages. There would be equipment shortages. Who gets ECMO if you system only has 12 machines?

I am not optimistic about quickly training people outside of medicine to become useful, but we have a lot of room for cross training and utilization within the field. Maybe we could use non -trained people to help with "orderly"type jobs. Heck, have some computer geeks follow around nursing staff and docs and they could do all the data entry on EPIC. Could save staff tons of time. Just quickly pass a safe harbor law so that if the non-medical computer geek writes down something stupid you can get sued for it.


Immediate retrospective research and setting up tests to answer research questions tomorrow for this flu will be just as interesting, and more relevant, than the 1918 flu which occurred in a different economy and different health system.

What will be interesting, as a research question, is whether or how much the precautions people are taking today--not touching face, washing hands--will affect the number of persons who contract the regular seasonal flu.

Are there other similar studies which can flow from this change in behavior and what data should we be gathering now?

How about a network analysis of the effect on the leisure market (travel, hotels, etc.) and substitution effects (purchases of large screen TVs and at home entertainment or teleconferencing systems)?

Will restaurants promote open air seating rather than eating indoors?

Will some hotel chain or cruise line send you a nose swab in advance of your trip so they can claim that all passengers had been cleared so you are safe to board to interact with other cleared passengers and crew in the bubble.

Will the movie Bubble Boy be replayed and become popular again?

Who will get naming rights to the next flu? It was unfair to call it the Spanish flu as it probably did not originate there, but, since we had had a war with Spain only 20 years earlier, perhaps it was easier to tag them with it. What about calling it the Wuhan or Xi Jinping flu? If you were threatened with having a flu named after you, you might be more diligent in preventing its spread.

But the important thing is we remove the posters that say "wash your hands."

That’s certainly stupid, and the order was rescinded immediately.

Even weirder that it was a career bureaucrat.

You mean someone that could be fired by the Trump Administration? Strange that they would place ensuring they keep their job ahead of serving the American people. Clearly, Trump needs to be re-elected to remove such self-serving lackeys from any position of responsibility.

"Christopher A. Santoro was appointed as the principal deputy chief immigration judge in May 2018"

But sure, if you think your job is to show up and undermine fair criticisms, you'll go with "career bureaucrat" rather than "recent appointee."

So, he could be fired at any time, it appears.

Surprise, surprise, surprise.

From 2009 to 2011, he served as special advisor for Enforcement and Removal Operations, Immigration and Customs Enforcement, Department of Homeland Security (DHS). From 2005 to 2009, Judge Santoro served in leadership roles in the Office of Inspection, Transportation Security Administration, DHS, including deputy director, senior advisor, and counsel. From 2001 to 2005, Judge Santoro was a trial attorney in the Criminal Section, Civil Rights Division, Department of Justice.

Try again

So, this means he was a man able to get appointed to leadership positions from which he could be fired by the president at any time under three administrations?

Sounds like exactly the sort of person that would think messages concerning hand washing would be a waste of time, and not correspond to the proper messaging that the current administration has tended to favor, even as recently as Monday, with Trump tweeting “So last year 37,000 Americans died from the common Flu. It averages between 27,000 and 70,000 per year. Nothing is shut down, life & the economy go on. At this moment there are 546 confirmed cases of CoronaVirus, with 22 deaths. Think about that!”

Typical bureaucrat - the signs were not authorized.

Actually, he was clearly responding to all that fearmongering and fake news.

Don't worry, after the U.S. gets its act together,useful information will be disseminated in a timely and efficient manner.

During the Spanish Flu my grandfather selflessly cared for a gravely ill friend, who managed to survive and gratefully repaid him -- with a racehorse.

Unfortunately the horse turned out to be not exactly Seabiscuit, so there was no lasting affect on my family, and no legendary racing saga for Hollywood or Laura Hillenbrand.

Never look a gift horse in the mouth (it might be a Trojan horse)?

Were fewer major league baseball players born in 1919 than in 1918 or 1920? That shouldn't be too hard for a sabermetrician to check at

The Spanish flu hit young adults hardest, those who were in their reproductive years. For children who lost one or both parents, it's not surprising they had difficulties in later years including health issues.

Inflammatory maternal immune activation (e.g. interleukin-6 cytokines) in response to pathogens has been shown to have adverse neurological effects on babies in utero. I'm not surprised that the Spanish Flu illustrates this on a large scale, but it is fascinating (and a bit worrying).

This whole premature birth reducing IQ thing is quite son was born 8 weeks early, has a officially tested IQ of 140 (tested to get into a gifted education program), using these numbers presented as "official science" he thus would have had an IQ of 160!! Perhaps it was good that he was born prematurely, as kids with 160 IQ's really struggle to fit into normal society!

Glad it worked out for you, but maybe environmental factors played a role too?

AlexT: "The 1918 influenza pandemic struck the United States with most ferocity in October of 1918 and then over the next four months killed more people than all the US combat deaths of the 20th century" - wrong. AlexT is channeling Barbara Tuchman, the sensationalist historian of the Black Death (her father was a prominent chess patron and banker too). In fact, historians of the sober kind say the Spanish flu killed not 10% of the USA's approximately 100M people around 1918, but rather (Wikipedia): "But a reassessment in 2018 [of the Spanish Flu] estimated the total to be about 17 million,[3] though this has been contested.[53] With a world population of 1800 to 1900 million,[54] these estimates correspond to between 1 and 6 percent of the population." Hence, 666k combat deaths for the USA in the 20th century would mean more people died of the Spanish flu in four months than this in the USA? Not likely, if you take even 100% of the American population infected (which is too high), and from the graph it shows a mere 100k at the peak, hence, unlikely this stat is realized.

I don't know if this has been mentioned, but my grandmother was 10 in 1917 and living in Gloucester, Massachusetts. Her favorite sister died of the flu. But what I'd like to mention is the fact that my grandmother was very superstitious, and always had a personal interpretation of tragedy. That's an aspect of human life that has largely been gotten over in this country I'm assuming. To make the point clearer, however many people were effected by the flu, if it hit your family harder than others then there had to be a reason, something you or someone in the family had done.

Coronavirus and pregnancy: What you need to know?
Experts simply don’t know very much about how coronavirus may affect pregnant women, but here’s what they do know

Comments for this post are closed