Wednesday, February 2, 2011

Runaway Trolley - Applied to Real-World Issues

The Runaway Trolley ethical choice thought experiment is a classic that has been tested on tens of thousands of people in different societies with remarkably consistent results. This may expose at least one basic human ethical value.



Click HERE to download a narrated PowerPoint Show that includes animated charts for the Runaway Trolley thought experiment. After the Runaway Trolley is explored, the charts continue and apply the ethical lesson to two real-world
issues: 1) Criminal Recidivism and 2) End of Life Issues. This posting covers the Runaway Trolley only. Subsequent postings in this series will cover Criminal Recidivism and End of Life Issues. The PowerPoint Show is based on a talk I gave to The Philosophy Club at The Villages, FL, on 04 February 2011. NOTE: The Powerpoint Show is now Narrated and plays and advances automatically after download to your computer.


The graphic illustrates the issue. A trolley has lost its brakes and is roaring down a hill. If nothing is done (FATE) the trolley will crash into the station at the foot of the hill, certainly killing at least five people.

There happens to be a junction and a siding. If that trolley can be switched over to the siding, the trolley will certainly stop safely in a pile of sand, and the five or more people on the trolley and in the station will not be killed.

YOU happen to be standing by the junction and see a switch that may be thrown to redirect the trolley from the Main Line to the Siding.

You are about to throw the switch (ACT) when you notice there is one person who happens to be sitting in the sand pile and that person will certainly be killed if the trolley is switched to the siding. What should you do?

NOTE: This type of thought experiment is intended to illustrate an ethical concept, so it is somewhat artificial. To play this game, you MUST ASSUME that all the facts stated are absolutely certain and, further, that all the people involved are total strangers and completely innocent. Therefore, you cannot avoid the ethical issue and suggest you could shout a warning to the guy on the siding - he is too far away to hear you. You cannot defer to a trolley company employee or another passer-by who happens to be at the junction - you are totally alone. You cannot phone the trolley company - time is short and you must let FATE take its toll or ACT and save several lives.

OK, now, what is THE RIGHT THING TO DO? Should you do nothing (FATE) or throw that switch (ACT)? Please decide now, before you read on.


The situation is similar to Part 1, the trolley is roaring down the track and five people will die when it certainly crashes into the station. But, this time, there is no siding. OY!

However, there is a footbridge that crosses over the trolley tracks, and, guess what, YOU happen to be on that footbridge. You see the trolley approaching the footbridge and you realize that if you do nothing (FATE) at least five people will die when the trolley crashes into the station.

Thinking quickly, you imagine yourself jumping from the footbridge onto the tracks in the path of the trolley. If you do so, you will certainly die but there is a chance your body will slow the trolley and save some or all the lives. Should you jump? You will die but five or more will live!

You notice that there is a very fat man standing on the other side of the footbridge, looking towards the station. He is totally unaware of the emergency, and there is no time to talk to him about it. He happens to be standing right above the trolley track and all it would take would be an easy nudge and he would tumble onto the tracks. He is so heavy he will certainly stop the trolley and save all the people, but, sadly, he will certainly die. He will die but five or more will live!

OK, now, what is THE RIGHT THING TO DO? Should you do nothing (FATE) or nudge that fat man down onto the tracks (ACT)? Please decide now, before you read on.


This problem, and versions suitable for "primitive" societies who do not know what a trolley is, has been posed to tens of thousands of people. I will tell you the remarkably consistent results further down in this posting.

Principle of Double Effect

But first, we need to learn about the Principle of Double Effect. It holds that:

You may take action which has bad side effects, but deliberately intending harm (even for good causes), is ethically wrong.
Thus, it is OK to act if your intent is to save many lives, even if, indirectly, some few lives are lost. This clearly applies to the Runaway Trolley Part 1 - the Siding Guy. Your INTENT is to save many people, and the Siding Guy's death, though clearly forseeable, is an UNintended side effect.

It is similar to the aircraft pilot whose plane is certain to crash. If he allows FATE to take its course, the airplane will crash into the center of a large metropolitan area, killing hundreds of people. If he ACTS he can divert the airplane to a less populated place, preferably a deserted area, but, if that choice is not available, he should crash into an area of single-family homes rather than apartment buildings.

Conversely, it is NOT OK to act if your intent is to kill someone, even if, indirectly, many lives will be saved. This clearly applies to the Runaway Trolley Part 2 - the Fat Man. Your INTENT is to kill the fat man, and the saving of the many lives on the trolley and in the station, is a side effect of a wrong intent.

(A strict Utilitarian might have trouble with that argument. What if killing one innocent person was certain to save 100 lives? 1,000? 1,000,000???)

This type of reasoning applies to an ethical thought experient called the Surgeon and the Stranger. A Surgeon at a hospital has five patients, each of whom needs a different organ transplant (heart, lung, etc.) or they will certainly die. There are absolutely no organs available from any normal source in time to save their lives.

Then a total Stranger checks into the hospital emergency room for a minor problem. The Surgeon learns that he is a drifter, with no relatives or friends, and no one in the world knows where he is or even cares. And, guess what, his blood type happens to match all five patients who need transplants. Your compassionate nurse suggests they fake the medical record of the Stranger to claim he died unexpectedly and they use his organs to save the five patients who need them desperately. "The five people who will die if we do not ACT have lived in our town for their whole lives. No one will miss this worthless Stranger. God must have sent that drifter to us for this purpose!" she says.

Well, as tempting as it is, the Principle of Double Effect says the Surgeon should let FATE take its course and not ACT. What do you think?

Statistical Results

The Runaway Trolley was presented to The Philosophy Club a year ago and everyone (100% of the 45 people in attendance) would ACT (throw the switch) to save five or more innocents despite the death of the Siding Guy. In the case of the Fat Man, 7% (three of the 45) would ACT (nudge him off the footbridge) while 93% would let FATE take its course.

A BBC poll found similar results, with 77% ACTing to save five or more innocents to the demise of the Siding Guy and only 27% ACTing to nudge the Fat Man to his demise, despite that opportunity to save five or more innocents.

NOTE: Subsequent postings in this series will extend this lesson to the real-world situations of: Criminal Recidivism and End of Life Issues.

Ira Glickstein


Howard Pattee said...

A great presentation! Do you think one C- and L-mind difference is the number n?


Ira Glickstein said...

Thanks Howard for downloading the PowerPoint Show HERE and going through all of it to the part about Criminal Recidivism.

For others reading this, the "n" Howard references is the number you'd put in the statement:

It is better to have n guilty go free than wrongly convict one innocent.

n varies from 1000 (Maimonides) to 100 (Franklin) to 10 (Blackstone) to 5 (Cardozo) to 1 (Voltaire). What is your n, Howard? I am with Voltaire.

And yes, I am sure L-minds would pick a higher value for n than C-minds. Do you agree?

Ira Glickstein

PS: I plan to have a new Topic on Criminal Recidivism next week.

Howard Pattee said...

Ira, I would say at least 2. (More depending on assumptions other than yours.)

I see your points about not publishing, and I agree. I only write now when I'm invited (but I'm still edited by peers).


Ira Glickstein said...

Well, Howard, with your n=2+ you are at least twice as compassionate as I. Please detail your "assumptions other than [Ira's].

My assumptions (in my PowerPoint charts) are based on criminal recidivism rates. It is established fact that more than 50% of those who were convicted of violent crimes, and served their sentences, and were released, were later caught and again convicted of another violent crime! That means, every time we release two convicts, we are condemning at least one of our innocent fellow citizens to violent injury. Probably more, because violent crimes usually have more than one victim, and criminals usually commit several crimes before they are caught.

Given 50%+ recidivism, among those who, having served their terms, are ten or twenty years older than they were when they committed their first violent crime, and thus (statistically) less dangerous, what does that say about those younger punks who are arrested, charged, and then get off on a technicality or due to evidence that is judged not to meet the "beyond reasonable doubt" standard. I think the guity who go free are more than 50% likely to commit another violent crime. What do you think?

Thus, perhaps the standard for "reasonable" (in beyond reasonable doubt) should be 51%? In civil (non-criminal) cases, the standard is by a preponderance of the evidence, which would be 51%. That seems backwards. If someone who is guilty gets off on a civil infraction and then commits another simiar infraction, no one is physically injured. But, when a violent criminal who is guilty gets off, and then commits another violet crime, one or more innocents are beaten, shot, or killed.

Ira Glickstein

PS: It is great that you are still invited to write (not just here in this Blog but in highly respected learned journals :^). Keep up the good work! It is an inspiration to me, having just turned 72, that I may not lose my marbles over the next decade or so. (Or, have I lost them already?)

Deardra MacDonald said...

Ira, What a great topic! I am amazed at how often I have used this concept of Act vs Fate for "small" every day situation in life! This has opened a Pandora's box for all of your reader to seriously analyze why and how they made their everyday choices...

Howard Pattee said...

My ethical principles begins with compassion or empathy (as expressed by the Golden Rule) but is moderated by the requirement of the overall consequences (i.e., if compassion for Bob results in added danger to Alice). This requires that you consider all the information available in each individual case, not just averages (e.g., recidivism) or the “preponderance of evidence.”

As Ira says: “NOTE: This type of thought experiment is intended to illustrate an ethical concept, so it is somewhat artificial. To play this game, you MUST ASSUME that all the facts stated are absolutely certain and, further, that all the people involved are total strangers and completely innocent. Therefore, you cannot avoid the ethical issue. . . ”

Well, these assumptions just avoid all the difficult ethical issues. You are left with a simplistic “utility function” like the greatest good for the greatest number, with “good” being undefined and “number” being the only measure of the value of a human. (All the statistics show is that people do not follow the assumptions entirely objectively. Ira’s dispassionate analysis is logically consistent.)


Ira Glickstein said...

Thanks Howard, you are really digging into my Powerpoint!

Howard writes: "My ethical principles begins with compassion or empathy (as expressed by the Golden Rule) but is moderated by the requirement of the overall consequences (i.e., if compassion for Bob results in added danger to Alice)..."

He is refering to two characters I used to illustrate End of Life issues.

Both Alice and Bob require expensive, public-funded (Medicare) medical procedures. In my example of "Quality-Adjusted Life Years" it turns out that Alice is entitled to her procedure but Bob gets only palliative care.

Howard excuses this decision on the basis that compassion for Bob (spending public money on his procedure) would endanger Alice, and all the rest of us in the public-funded medical plan, by depleting resources to the point the plan was not sustainable. In President Obama's words: "I mean, the chronically ill and those toward the end of their lives are accounting for potentially 80 percent of the total health care bill out here."

Please let me know, Howard, if I mis-represented your view.

I do not fully understand the rest of Howard's comment. I seems to me that the best we can do is use all the information available in each individual case, and plug it into some preset decision matrix. We will never know everything with certainty. Alice may, like Obama's grandmother, die a couple weeks after her public-funded procedure, and Bob, had he received his procedure, might have lived on for 20 years. All we can do is use the best available medical experience and statistics.

Ira Glickstein

joel said...

There are a couple of things that bother me about this thought experiment. The first is that I don't think it should be called a thought experiment in the Maxwellian or Einsteinian sense. An experiment should contain all the elements necessary for its operation. The observer is only that. The observer could just as well be a machine. In the hypothetical called the runaway trolly, the "observer" or "decider" is himself or herself the subject of the experiment. I believe we should we should use the word "hypothetical."

If one concedes the above, the details of the scenario and the means by which it is presented become important. For example, Ira proposed the scenario at our meeting with an animated powerpoint presentation. (He did a great job.) One would not be terribly surprised if the reaction of the test subjects (audience) varied with the goriness of the images, the amount of time allowed to absorb the situation and the amount of time allowed for consideration and response. Even the use of the word "fat" to describe the trolley-stopping human may have an impact of some people. Thus the use of this so-called "thought experiment" as a tool to determine the universality of ethical concepts runs into cross-cultural differences and effects of small variables; a sort of butterfly effect in the human brain.

Ira Glickstein said...

Thanks for your kind words, Joel. As I presented the "fat man" scenario at the Philo Club I watched a couple of our most portly attendees to see if their personal physiques would affect their weighty votes!

I agree "hypothetical" (or perhaps "scenario") is a better term, but my Google search did not find "hypothetical" used for the Runaway Trolley type of mental exercise. Wikipeda has an entry for Thought Experiment, but not "hypothetical" in that sense. Their Trolley Problem first line says: "The trolley problem is a thought experiment in ethics, first introduced by Philippa Foot, ... [It] has been a significant feature in the fields of cognitive science and, more recently, of neuroethics."

It appears the first use of Gedankenexperiment (a German/Latin construction for thought-experiment), around 1812 was by Hans Christian ├śrsted, who also used the pure German equivalent Gedankenversuch in 1820. According to Wikipedia, the original usage was quite broad, but, "Much later, Ernst Mach used the term Gedankenexperiment in a different way to exclusively denote the imaginary conduct of a real experiment that would be subsequently performed as a real physical experiment by his students (thus the contrast between physical and mental experimentation) with Mach asking his students to provide him with explanations whenever it happened that the results from their subsequent, real, physical experiment had differed from those of their prior, imaginary experiment."

Of course Einstein's most famous EPR gedankenexperiment, posed, along with Podulsky and Rosen, in 1935, was not actually done until after his death.

I guess a despot could run the runaway trolley as a real experiment, to see what real people would do if they had actual control of the fate of the people on the trolley and those in the station as opposed that of the siding guy or the fat man. If that was done, I suspect most subjects would freeze and let FATE take its course rather than ACT.

As for your distinction between observer and decider do you think the results would be substantially different if I said "John ACTed in the case of the siding guy but Joe let FATE take its course in the fat man situation. Was John acting ethically? Joe?

When attendees at our Philo Club (or in the BBC poll) voted, I suspect they were more in the observer role because they had time to make up their minds, and they were asked to assume the facts were certain. An actual decider would have mere seconds, far less certainty, and more involvement and even legal jeopardy when his actions were reported to the authorities.

Ira Glickstein

Anonymous said...

I learn so much from your presentation at the philosophy club. One of my five books is about lawyers. I am glad you mentioned about clever lawyers and failure in the justice department. Maimonides, Ben Franklin, Blackstone, Cardozo are all new to me in what they said about letting go of the guilty.
M. Hwa Chan