Bryan Caplan  

Ebola Bet

Fixed Costs and Open Borders... Preferences in The Warriors...
Mainstream scientists assure us that Ebola poses very little threat to Americans; unless you're a health worker who cares for the infected, Ebola is almost impossible to catch in a rich, modern society.  Yet many populists and borderline conspiracy theorists are convinced that the experts are seriously understating the danger.  In their contrarian opinion, we desperately need to close the border now.

Fortunately, this is an easy argument to put to a bet.  My tentative offer: $100 says that less than 300 people will die of Ebola within the fifty United States by January 1, 2018.  I'm willing to switch to "Unless the U.S. changes its Ebola-related policies, $100 says that less than 300 people will die of Ebola within the fifty United States by January 1, 2018," but then we'd have to carefully define what policy changes count.

I will make this bet with up to five individuals with sufficient reputation to make payment likely if they lose.  I'm also happy to entertain alternative bets.  Propose them in the comments or email me directly.

Comments and Sharing

COMMENTS (39 to date)
Neil writes:

Shouldn't you make a bet with a larger death count and 1000:1 odds or something? It's certainly true that it's highly, highly unlikely this amounts to much, but it's those dire, worst-case scenarios that the populists and conspiracy theorists are interested in.

Arthur B. writes:

Will take it at 1000:1 (i.e. my 10c against your $100).

James Miller writes:

Ebola's threat to America comes from the small chance that it will kill a huge number of Americans. It might well be that there is less than a 50% chance that Ebola kills >299 Americans, but still more than a 1% chance that Ebola kills more than 10,000 Americans.

I will bet you $30 vs $3,000. You win $30 if by January 1, 2018 Ebola has killed

I offer you this bet for 48 hours.

James Miller
Associate Professor of Economics, Smith College

Duncan writes:

I understand betting markets can be great predictors, but Bryans bets always seem too small. $100 or even $500 I'm sure is not really that much money to him.

Now James Millers bet of a $3000 loss to Bryan would, I expect, be much more noticeable to him.

James Miller writes:

Sorry the full bet didn't get posted.

You win $30 if by January 1, 2018 Ebola has killed less than ten thousand people in the United States. I win $3,000 if by January 1, 2018 Ebola has killed ten thousand or more people in the United States. To avoid bad publicity if I win pay the money to one of the top Givewell charities. If you win I will pay you directly. I offer you this bet for 48 hours.

Arthur B. writes:

Just a rejoinder to what James is saying:

I've seen many people somewhat condescendingly explaining that fear of Ebola is ridiculous since the odds of catching it are very small.

That's assuming that people only care about their own survival. Ebola does carry a small probability of wiping out a large portion of humanity in a particularly gruesome way.

There's also probably a lot we don't know about Ebola. This calls for extreme caution, even if to the best of our current knowledge, a worldwide epidemic seems very unlikely.

David Collum writes:

It is marvelous that an economist is willing to make such a bet. Unfortunately, such an existential risk of admittedly unknown probability is worthy of more consideration.

As chair of the Department of Chemistry and Chemical Biology I have no more direct insight into this disease than you, Professor Kaplan. I do, however, appear to have a greater respect for what it is I do not yet understand. I also have access to biochemists, and a non-trivial number are not as confident as you. Try reading The Great Influenza. You will quickly realize that not only did 100 million people die worldwide in six months but that modern medecine would not have saved them.

The doubling time for Ebola is running around 3 weeks. Maybe it has already turned down. Nonetheless, economists are as qualified as any group to understand the miracles of compounding.

I will not take your bet because I can either lose the bet or win the bet; either way I lose. To call us conspiracy theorists, however, illustrates the breadth of your imagination. The term conspiracy theorist is designed to shut down open debate. Sorry. It hasn't worked on me. The term is repugnant.

David B. Collum

Melissa writes:

[Comment removed pending confirmation of email address and for foul language. Email the to discuss restoring your comment privileges. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Lemmy caution writes:

A good bet would be less than 300 deaths they pay you 300-x ; more than 300 deaths you pay them x-300 ; where x is the number of Ebola deaths. What number would you be willing to substitute for 300 before you would take this bet? That would be the true freak out equivalent point. My guess is that it is pretty high.

JKB writes:

Seems to me you should hedge against a more virulent mutation since this version is getting a better population run.

But also, Walter Russell Mead quoted this bit from the Guardian. The vectors may be different than most are looking at with all the talk of West Africans traveling. The Westerners acting foolish seem to be significant risk as well. Are the travel bans being promoted for them?

Samura believes sexual promiscuity among westerners could play a role in the virus’s spread abroad. Almost immediately after the outbreak was reported in March, Liberia’s health minister warned people to stop having sex because the virus was spread via bodily fluids as well as kissing.“I saw westerners in nightclubs, on beaches, guys picking up prostitutes,” he said. “Westerners who ought to know better are going to nightclubs and partying and dancing. It beggars belief. It’s scary.”
Mark Bahner writes:
Try reading The Great Influenza. You will quickly realize that not only did 100 million people die worldwide in six months but that modern medicine would not have saved them.

Oh, c'mon.

Just look at this photo of an "influenza hospital" at Fort Riley KS, from Wikipedia:

"Influenza hospital" in the 1918 epidemic

Even I could do better than that.

I could probably name several dozen technology-based reasons why flu will never kill 500,000+ people in the U.S., such HEPA filters, UV lighting, and positive pressure respirators. Look at what workers treating Ebola victims look like, and then look at what the nurses (and patients) at Fort Riley look like.

James writes:

Can anyone else corroborate David Collum's claim that modern medicine would not have saved the people who died of the 1918 flu?

David Collum writes:

Sent in a response that is either delayed or missed somehow.

The victims of the 1918 did not die of secondary infections and standard flu dehydrations (as I certainly believed). They died of what were called cytokine storms--massive immune responses that were most lethal for healthy people in their prime. Their symptoms were gas build up under the skin, making them sound like Rice Crispies when moved. Brain hemorrhages were very common. Asymptomatic patients could go to sleep in the evening and not regain consciousness. That is what is so nasty about systemic viral infections.

Andujar Cedeno writes:

Death isn't the only permanent consequence of Ebola. For example, if someone has a business in the travel industry or dependent on travel then the Ebola epidemic represents a clear danger to their livelihood. The threat is even greater to young companies with high debt ratios and still building revenue streams. How many businesses will Ebola bankrupt in the United States?

Troy Barry writes:

I'll take the bet (first form), not because I have any particular expertise or strong concern about ebola. (I do not believe closing US borders is justified by the threat, and wouldn't even if we knew 300 resultant deaths were a certainty.) But I am sceptical of some of the comforting assumptions of your mainstream scientists, namely:
- Human living conditions in West Africa and the USA are assumed to be so utterly different that they entirely change the disease's infectiousness. I'm not so confident that the African's lives are unsanitary or that American lives are impeccably sanitary.
- I suspect an element of downplaying for reassurance in the experts' description of the risk of airborne contagion.
- Symptons should manifest in infected carriers between Africa and the USA, enabling quarantine on arrival. That seems to overrate the effectiveness of border controls. How many people cross your borders per day - a million or so? There's going to be a lot of misses in the billion events before 2018, and there's some chance that some of those misses will overlap with ebola carrying arrivals.
- Related to that, a trip from West Africa to the USA might be a reasonably long trip, but to reach Miami from Conakry by 2018 the disease only needs to spread an average 10km per day, and the Atlantic has never proven a great barrier to infectious diseases. Will symptoms manifest before a carrier travels from Toronto, Havana or Panama?

You accept the risk that ebola transforms to a more infectious or less detectable form (although I don't consider it a significant factor so that's negotiable). The $100 is US currency so I accept the exchange rate risks. (My meagre wealth is held in AUD.) If the US becomes an ebola hosptial, deliberately importing ebola carriers e.g. for treatment, any known carriers deliberately entering are excluded from the count but resultant accidental deaths are not. Health workers count as people, but are likewise excluded if they are known by the authorities to be infected when entering the USA.

I acknowledge my reputation is insufficient to give you confidence in repayment, therefore I propose to transfer $100 to you on your acceptance. If you win the bet, you need never repay it. If you lose the bet you transfer me $251 in January 2018. (Or suggest your own estimate for the future value of the 2x$100 - which would be worth hearing in itself. :)

Mike Lorenz writes:

Bryan - I'll take that bet. Figure out a way to determine my reliability as a counterparty.

I hope you win.

Mark Bahner writes:
Can anyone else corroborate David Collum's claim that modern medicine would not have saved the people who died of the 1918 flu?

I think it's complete nonsense. Recent research indicates very strongly that most deaths in the 1918 pandemic were from secondary bacterial pneumonia infections:

Bacterial pneumonia a significant cause of death in 1918 flu pandemic

To say that modern medicine couldn't save people dying from bacterial pneumonia is nonsense. People seem to have little appreciation for how tremendously primitive medicine and medical technology was 100 years ago.

Todd Kreider writes:

Always great to see Mark Bahner posting!

Mike writes:

Ebola:Africa::WMDs:Iraq. It's the only way you can justify the US response.
* Restrict inbound flights from infected countries? No.
* Send medical aid? No.
* Import "infected" patients? Yes.
* Send 5,000 troops? YES!

Americans didn't fall for Kony 2012, and they didn't fall for the kidnapped children of Nigeria, so they had to invent a new, scary story to somehow justify sending the military to Africa. Nobody really believes you can stop Ebola with bullets, but it's too late for that, the troops are already there.

Nick writes:

An interesting variation is compared to the 1918 flu. Population in 1918 was about 100 million. Estimates of American dead from the flu is 675,000, or about 0.67% of the population. (Also note the scary one year decrease in population during that year in the census data). As a percent of population, that would be about 2 million Americans dead today. What odds would someone take at and over/under level of 2,000,000 Americans dead?

I think that creates a prediction market that tries to value some important information, namely, is Ebola going to be a significant problem in the US? 300 people seems like too small a number to answer that; obviously if fewer than 300 people die then Ebola will not have been (in hindsight) a big problem, but if more than 300 people die it might or might not. 301 people dying, while a tragedy, would satisfy the conditions of the bet but would not mean that there was a significant problem, while 5 million would both satisfy the bet and be a significant problem. A good prediction market should be able to tell the difference.

lemmy caution writes:

The Ebola risk is a small risk of a high number of deaths.

Consider the bet

less than C number of Ebola deaths (x):

I pay you C-x

more than C number of deaths:

you pay me x-C

What C would you be willing to accept under these terms. My bet is that it would be pretty high. Who would risk their life savings?

RPLong writes:

I'm surprised so many people are objecting to the terms of Prof. Caplan's bet. He's not trying to monetize the value of each Ebola death, he's placing a premium on his own level of confidence.

I won't take the bet, because I believe Caplan will win. That said, I was surprised that a 2nd nurse was diagnosed, so my confidence is less this week than it was last week.

ColoComment writes:

Andujar Cedeno writes:

I'm with Andujar. Given our state of the art medical processes, I suspect that the number of actual deaths from Ebola may be minimized. But death is not the only consequence: the economic effects of American victims may be even more telling.

Remember when markets and business were closed after 9/11? Airlines did not fly. Business came to a halt. The economic life of our country was interrupted to the tune of untold billions of dollars of business not transacted.

Even if death does not occur, all of the Ebola victims' personal contacts must be found and quarantined: those people do not go to work, those students do not go to school, those bus drivers do not drive, those waiters do not serve, those lawyers do not make wills, and so on.

It is not the number of deaths with which we should be concerned, it is the spiderweb of effects of potential contagion.

MikeDC writes:

I would take the bet with the proviso that it be cancelled in the event significant travel restrictions are imposed, since that would render the bet moot on settling the underlying policy question.

I tend to think >300 people will die in the US by 2018 because:
1. The epidemic does not seem to be abating in Liberia, Sierra Leone and Guinea. Rather, WHO officials say it's growing quickly, and reappearing in areas that were thought to have been fixed.

That means, at best, the threat of the virus leaving those countries is likely to continue to grow and be present over the next several months or possibly years.

2. The level of training and mission of the US military personnel being sent there seems to be very unclear.

3. And, we aren't imposing any sort of quarantine that I can tell.

My guess is that this causes more cases to trickle in to the US over the coming months and years. One of these will be "unlucky" and expose a number of people here who will get it.

Michael Barry writes:

This (BC's) post typifies the "anyone not for open borders is a Neanderthal" approach to this issue. No doubt there are "populists and borderline [pun intended?] conspiracy theorists" who want to "close the border now." I think all most of us are saying is, let's just not give visas to Liberians who want to come to the US to visit their girlfriends (or who have similarly un-compelling reasons for US bound travel).
I'm guessing that -- unless Ebola spreads to a Latin American country -- less than 30 Americans will die of it, even if we do let Liberians visit their US girlfriends. But the cost vs. benefit (to Americans -- not really sure why I need to worry about the cost/benefit to Liberians-visiting-their-girlfriends) of unrestricted Liberia-to-US travel isn't persuading me.
Thanks for the condescension,
Michael Barry

David B. Collum writes:

If Ebola hits big cities in China it would be bad too, but I bet they would be decisive. For me, the risks jump when Ebola gets to India--a petri dish of 900 million people, crowding, and easily overwhelmed healthcare system. It also then goes Out of Africa.

BTW-The link to the recent studies on the flu was a good one. I made the mistake of making a refutable argument (unknown at the time and still maybe debatable) that wasn't necessary. Ebola is a killer even in the first world. It is our containment strategy that may or may not differ.

I have been exchanging ideas with a prominent global macroeconomist on the economic impact of even low Ebola counts in the US that I believe will result in his analysis. (Don't wish to divulge his name but y'all probably know him.) I will be writing about it in December in the context of an overview of the year's events.

In any case, this is a story of great interest and importance even if it fizzles relative to the worst case scenario. I probably should have let my pulse settle down on my first post but the term "conspiracy theorist" fries me pretty fast. It is ad hominem at its finest.

Carl writes:
Yet many populists and borderline conspiracy theorists are convinced that the experts are seriously understating the danger.

Mr Collum, this does not necessarily apply to you, does it? Unless you identfiy as a "populist" or a "borderline conspiracy theorist", of course. You don't. So it doesn't!

Colm Barry writes:

well, I'm not the betting type. However: while the Netherlands ALWAYS quarantine everyone who comes to a hospital, then test them, then only, if proven "clean" release them to the general ward, most other countries do the opposite: they admit everyone, whether he/she may carry MRSA or Ebola, to infect everyone else and only as an afterthought then treat 70% of hospital-induced infections as a result. Mostly these are "harmless". But do this with just ONE Ebola patient and you'll have lost your bet. The Netherlands have but 1% of hospital-induced infections ...

David B. Collum writes:

I am, indeed, a borderline conspiracy theorist. I believe that banks rigged libor (check), that Paulson lied about TARP (check), that Saudis push oil prices down for geopolitical reasons (check), that one Juncker said you had to lie that was not at lie. I believe in the notion of "deep politics"--the politics below the surface.

Mark Bahner writes:
Mr Collum, this does not necessarily apply to you, does it? Unless you identfiy as a "populist" or a "borderline conspiracy theorist", of course.

One needn't identify as either to find the characterizations unnecessary and ad hominem. (Though in my opinion an order of magnitude or more less so than "#LibertariansForEbola".)

It's just not necessary to go farther than, "I think they're wrong. Here's why." Or, "And here's a test."

Best wishes*,

*P.S. To all. Even those who disagree with me...and therefore are probably wrong. ;-)

David B. Collum writes:

Well said, Mark.

James H. Stein writes:

I offer to accept your bet. (I hope I lose.) Having insufficient reputation (in your circles) to guarantee my payment, I'll let you hold my $100 until the bet's outcome is clear.

On another topic: I observe that the "factoid" that you cannot catch Ebola from someone who is asymptomatic is claimed by WHO (who ought to know) and is endlessly repeated by journalists, but supporting reasons are never given. Who can explain why the factoid is a fact?

John B writes:

[Comment removed pending confirmation of email address. Email the to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

khodge writes:

I'm not sure I understand the this about Ebola or is this about the value of $100.00 in 3 years?

Granite26 writes:

My concern about the bet is that it doesn't specify who is doing the dying. Right now we're not catching a lot of international travelers with ebola, nor or we seeing a lot. That's something that could change at any time.

Assume an infected person hops on a plane and starts showing symptoms (serious symptoms?) in flight. Is the US going to turn him away, or immediately throw him in quarantine? Does he count as dying in America?

What about the others on the plane? 300+ people possibly infected, even at a low infection rate. Even if they are immediately quarantined on hitting the ground and have 0 risk of further infecting others, there's still a serious risk of them dying in America.

Finally, we have a large number of international aid missions, including apparently US troops. We've shown a general willingness to fly US citizens back for treatment.

All told, and with 0 facts, I'd guess that the mortality of ebola for properly treated US patients to be 30-50%. In that case, it would take around 450 people flying to the country with ebola and each infecting .5 people in transit in order to reach your 300 people.

If you were to modify the terms to say Americans dying of the disease who caught the disease in the US, I'd say you're probably very safe. Given that we seem comfortable (and rightly so) leaving the borders open, I think we're going to see a lot of ebola deaths on US soil (and that's ok).

Carl Pham writes:

I observe that the "factoid" that you cannot catch Ebola from someone who is asymptomatic is claimed by WHO (who ought to know) and is endlessly repeated by journalists, but supporting reasons are never given. Who can explain why the factoid is a fact?

Viruses are, contrary to popular imagination, rather fragile structures, which fully retain their functionality only when immersed in an ideal medium -- water solution, of a certain pH and ionic strength. As soon as the medium changes, the viral particles begin to degrade and lose functionality.

The most important issue is dehydration. If the viral particle is not fully immersed in water, the proteins in its outer coat begin to unravel. So anything that dries out a virus tends to ruin it. (In fact, a quick way to make a vaccine is to partially dry out the virus, this renders it nonfunctional without destroying the bits and chunks of protein in its outer coat that can be learned to be targeted by the immune system.)

Viruses differ tremendously in how resistant they are to dehydration, which is their inevitable fate as soon as they are ejected from inside a living body. Some, such as the coronaviruses that cause colds, are quite resistant. For that reason, they can readily be transmitted through the air (in tiny droplets) and remain functional for minutes to hours on surfaces. (The best surfaces for preserving viral function are hard hydrophobic surfaces -- polished dry stainless steel is best -- because these do not absorb the water surrounding them.) The HIV virus is poorly resistant to dehydration, so you pretty much have to have direct blood-to-blood transmission. The Ebola virus is in between, not as resistant as the cold or influenza viruses, more so than AIDS.

So, generally, if you are not actually emitting bodily fluids containing the Ebola virus -- not exhibiting uncontrolled bleeding and diarrhea, then the viruses you shed are few and tend to become nonfunctional quickly. Hence, it is difficult -- not impossible, just difficult -- to catch.

Viral transmission is always a probabilistic thing. You can't say it's "impossible" or "possible," it's not binary. You can't even say it's "possible" or "impossible" by a given route -- it's not a question of whether the virus has wings or not. Whether viral transmission occurs is a combination of how resistant to dehydration the virus is, in what bodily fluids the virus is shed (which depends on what tissue it infects), whether it provokes emission of those fluids (the way the cold viruses provoke sneezing and runny noses), and in what strange concatenation of microscopic events provides a more or less moist transmission path from one host to another. There are no black-and-white answers about transmission, there are just various shades of gray through which experts try to draw the bright lines that ignorant yoyos and journalists (but I repeat myself) demand.

vidyohs blanco writes:

Referring to the stipulations on who would be acceptable to bet against: Considering your quibble, let's suggest that you put your money up with a neutral party as do the ones with whom you bet, that way every one will be assured there will be no welchers. And, one further question. If one of our soldiers gets Ebola in Africa and dies there does it count against your proposed total of 300?

Rob Larson writes:

Carl Pham - thank you for that explanation. Most clear and informative explanation of virus transmission I've ever read. (Of course, I'm taking it on faith as being accurate, since I don't have the expertise to judge.) Thanks for taking the time to inform the rest of us.

David Boudeau writes:

And, what do the unlucky ones who die from Ebola get? After all, it's a small sacrifice to make when compared to the horror of preventing people from the ravaged country to come here at will.

Comments for this entry have been closed
Return to top