Deciding abortions by the toss of a coin?

Author: Michael A. Lewis

I was recently listening to public radio one morning. I live in the U.S. and in the last few days there’s been a lot of attention being paid to a Texas state Senator named Wendy Davis. Due to procedural rules in the Texas legislature that I don’t need to go into, Davis was able to stop, for the time being, an abortion bill that in her view would have greatly decreased Texas women’s access to abortion. She did this by extending “debate” on the bill (what we call a filibuster in the U.S.) by talking for over 10 hours - she didn’t even take a break to “go potty” as my daughter used to say.

Just a placeholder image

Woman holding sonogram

Courtesy of Creatas Images:

The Davis episode is the latest skirmish in an ongoing battle in the U.S. between pro-choice and anti-abortion forces (the term those who are against abortion prefer is “pro-life”). The pro-choice side argues that women with consultation from their physicians should decide whether or not to have an abortion, not politicians. The anti-choice side argues that abortion amounts to the murder of a child and, therefore, there should be laws to stop such murder from taking place. Yet based on what I’ve read in newspapers and heard on talk shows, some in the anti-choice group would allow abortions under some very exceptional circumstances.

The most common exception I’ve heard is when the health or life of the mother is at stake. In other words, according to some anti-choice folks, if forcing a mother to carry a fetus (they’d probably say “baby”) to term could result in severe negative health consequences or the death of the mother, then the law should allow an abortion. But abortion in all other circumstances would be disallowed.

Now, in full disclosure I’m pro-choice. I’m so pro-choice, in fact, that some of my pro-choice friends are uncomfortable with just how pro-choice I am. But I won’t go into those details in this post. I’m also a social worker/sociologist who specializes in quantitative methods. The quantitative methodologist in me has always wondered something about those who are anti-choice but would allow for some of the exceptions discussed earlier.

If abortion is the murder of a child, why is it okay to murder a child to protect the life or health of the mother? That is, if we’re faced with a situation where forcing a woman to carry a fetus to term would (or would likely) result in that mother’s death but an abortion would result in the “child’s” death, why is it right to kill the child to save the mother? From the point of view of someone who is anti-choice, it would seem that an arguably better way to resolve this dilemma would be to draw on the laws of probability.

When faced with the situation spelled out in the last paragraph, suppose we did the following. Assign the mother the number 1 and the fetus the number 0. Next we’d use the random digit generator in a computer to simulate a coin toss (to avoid the possibility of someone being able to manipulate a real coin by tossing it in a certain way) by randomly choosing either 1 or 0. If the number 0 is chosen, the abortion occurs, resulting in the death of the child. If the number 1 is chosen the mother is forced to carry the baby to term, resulting in her death.

I’m personally horrified by this proposal and suspect other pro-choice proponents are too. But are anti-choice folk who support exceptions, such as those discussed earlier, horrified by it? From their point of view, wouldn’t randomly deciding who dies in such situations be more just than automatically killing a child to save their mother?

My gut tells me that many anti-choice proponents of the life/health of mother exception wouldn’t like deciding who dies in these situations by effectively flipping a coin. If I’m right, then there is an obvious question - why not?  Could it be that their preference for sparing the mother’s life and health over the baby’s is an implicit endorsement, at least in part, of the pro-choice position?

Bookmark and Share

Comment on this article

Submit your comment
  1. Image of unique ID


Perry Neheum

Religious beliefs should NOT in this time propel abortion discussions.


"Cutting Commentary"

Abortion foes never want
to admit the fact they won’t confront:
Those unviable tissues that they treasure
Are not babies by any measure.

Yet righteous guys, who never bear
a mote of life, will fiercely swear,
It's against God’s will to terminate
cells gals alone can germinate.

As long as dogma directs the issue
of what to make of useless tissue,
Womens’ rights will be repressed
by those who think they know what’s best.

-- By Kenneth S. "Ken" Spalding

reply to this comment


I agree with Adam.  I read Significance for its penetrating statistical insights, not politically charged op-eds.  There are literally millions of blogs for this kind of content.

reply to this comment


As a subscriber, I'm ashamed that Significance allowed this to be printed. The scenario is like those that teenagers argue late at night. Here's effectively the same thing: the two people Michael A. Lewis cares for the most are in a car with him, which crashes. They are hurt. He can get out and pull one of them with him before the car explodes, so he should mentally flip a coin. There's no statistical thinking involved in either case.

In the actual world, the argument about the mother's life comes from situations where pregnancy would cause the mother to die, which would also cause the unborn child to not reach term. People argue one against two. Other people argue that a fetus can never be killed, but certain medical procedures such as chemotherapy or removal of ectopic attachment save the mother's life, and has an unavoidable unintended effect of killing the embryo.
People who argue that point would not be "negotiating the price", a quote attributed to Winston Churchill.

reply to this comment


All you people who are offended by this article and find it stupid, I think are not quite getting the point. Yes, the purpose IS to show how absurd flipping a coin to decide whether to get an abortion can be, if you are pro-life. Pro-choice people will most likely also find it absurd to flip a coin to decide whether to get an abortion.

The point is to get to the gut of what pro-lifers base their pov on--that all life is precious, and to try to formulate a probablistic experiment to see how/why they would choose the mother over the child if the mother's life is in danger. The author IS going back to statistics, and IS trying to use logic to tackly this tricky scenario. So if you find this stupid and offensive, it would be nice if you'd respond with some equally "statistical" or "mathematic" or "intelligent" analysis, in order to elucidate our minds.

Thank you.

reply to this comment

Kyle Simmons

In principle then, pro-lifers do condone abortion, and "once you agree in principle, it's all about negotiating the price". Don't quite remember who said that.

reply to this comment


This is too stupid to be serious.  If it is a spoof, it lacks the prerequisite of humor. 

reply to this comment

Julian Champkin

William: Is your problem with the morality of abortion, or with Lewis' assertion that coin tosses are, in real life, a good way of generating random numbers?
"If the author were able to model the physics of coin tosses he'd realize that coin flips are fully deterministic. Meaning he'd have to explicitly pick whom to murder. " But neither the author, nor anyone else (except Persi Diaconis - see Significance March 2006) CAN model the physics of a single real-life coin toss to determine which way it comes down: can YOU flip a coin so that it always comes down heads?  So he is not remotely explicitly choosing whom to kill. Surely coin-tosses are the most standard way statisticians illustrate random two-way choices?  
Can't a statistics magazine mention coin-tosses? Lots of statistical authors do....

reply to this comment


I think they'd probably prefer praying

reply to this comment

William M Briggs

Lewis's article has nothing to do with coin flips, Julian. It is an absurd excuse to push a political position. If you like, I'd be glad to offer a full-length rebuttal. Plus, Persi has certainly modeled the physics of coin flips and often lectures on it. And as a matter of fact I CAN flip a coin so that it lands the way I want it. Like Persi, I have long dabbled in amatuer magic. So, no, coin flips are not good "random" number generators.

reply to this comment

William M Briggs

What an appallingly stupid article. No, I don't mean the immoral, murderous position so gleefully supported by its evil author. I mean the asinine attempt to tie the killing of the lives inside not-to-be mothers to probability, in a journal purpotedly about statistics. What makes it worst of all is the concept of probability is wrong. Use a computer to "simulate a coin toss"? If the author were able to model the physics of coin tosses he'd realize that coin flips are fully deterministic. Meaning he'd have to explicitly pick whom to murder. So much for his scheme. Come on, editors. Show some pride and common sense. Eschew blatant and ridiculous screeds like this. Go back to statistics.

reply to this comment

Skip to Main Site Navigation / Login

Site Search Form

Site Search