Lately, social media has been flooded with all sorts of "facts" about abortion. This one in particular caught my attention:

42% of all yearly deaths in the world are from abortion

42% of yearly deaths in the world are from abortion

A worldwide crisis

I don't want this to degrade into an argument about what "death" is, so for the sake of the clarity assume abortion is death.

Is this statistic correct?

  • 33
    But is a natural miscarriage death?
    – Henry
    Commented Jul 23, 2013 at 6:27
  • 2
    A more interesting statistic might be the worldwide abortion rate versus the rate for other forms of non-natural death, such as murder, war, genocide, suicide, etc. For the first point, I found this imperfectly-formatted data, and was surprised that it seems to come out to well under 500,000. The total must be well under 42 million. Depressing. Commented Jul 23, 2013 at 9:32
  • 5
    The only reason that claim even remotely works is by a very narrow and warped definition of what does and doesn't constitute a "death" in order to slant the statistics, and as the answers point out the only way to get those numbers is to use multiple definitions on each percentile. As such I'd call the graph completely and utterly -- and deliberately -- misleading, and therefore untrue. Commented Jul 24, 2013 at 8:58
  • 15
    If true, this statistic is a pretty strong argument in favour of permitting abortion; overpopulation would be rampant otherwise.
    – gerrit
    Commented Jul 24, 2013 at 10:37
  • 5
    @gerrit: no one who (reasonably) opposes abortion does so on the basis that we need more people. Many secular arguments against abortion rely on the fetal personhood. If fetuses are persons, you can't kill them, even if there is rampant overpopulation. Overpopulation might be an argument in favor of permitting abortion, but not a good argument against major secular arguments against abortion. Commented Aug 21, 2013 at 5:15

3 Answers 3


No, the picture is not accurate. No, “42% of yearly deaths in the world” are not from abortion.

As Avi’s answer shows, the raw numbers are about right, albeit with a big uncertainty attached.

However, the percentages are wrong because the 42% implies that the “total number of deaths” = “number of deaths after birth” + “number of abortions”.

This is inaccurate. There is at least one additional factor:

As many as 50% of all pregnancies are miscarried before coming to full term – most (about 60%) due to chromosomal aberrations. If we count pre-birth deaths in our statistic, then we also need to count those miscarriages.

According to the CIA World Factbook, the birth rate in 2012 was 19.15 per 1000, that’s 135.97 million births, and consequently about 135.97 million miscarriages.

If we add the numbers, we find that abortions account for 20% yearly deaths in the world, not 42% as claimed (and this isn’t even accounting for the fact that of those 20%, some would have miscarried anyway).

Of course this is assuming that we count prenatal deaths as “deaths” at all, and as can be seen from the statistics on mortality rate, this is not normally done. The most we can honestly say is that “there are 0.72 as many abortions as there are (postnatal) deaths (= 72%)”. But presenting these 42% as part of the overall deaths is wrong, as I’ve shown above. In statistics, presentation matters.

Here’s a summary:

Comparison of a correct plot with an incorrect plot of the data

(The numbers differ slightly from those in Avi’s answer since I’m using the “42%” number from the question to calculate abortions but the principle doesn’t change.)

  • 6
    This is messy, as we are already sinking into the quagmire of differing definitions. It seems the claim and the first two answers are agreeing on the facts, but disagreeing on the way of presenting them.
    – Oddthinking
    Commented Jul 23, 2013 at 9:30
  • 14
    @Oddthinking They’re agreeing on the absolute numbers, not on the relative numbers (this is also a fact). I agree that the definitions are messy but unless there’s compelling evidence for treating abortions as deaths and not miscarriages, and given the fact that the normal definition of death does not include any prenatal deaths, I think it’s safe and consistent to say that the 48% number is wrong. Commented Jul 23, 2013 at 9:39
  • 4
    For the record, I think it does make substantially more sense to include miscarriages if you're including abortions. I just judged that as outside the scope of this question.
    – Publius
    Commented Jul 23, 2013 at 14:33
  • 1
    50% is a pretty conservative figure for miscarriages. Only about 30% of fertilised eggs go to term without any intervention from abortion. Commented Jul 23, 2013 at 14:38
  • 7
    A minor point, but your statement about "there are 42% as many abortions as there are (postnatal) deaths" also isn't accurate as you aren't taking the abortions out of it. The accurate statement would be that there are 72% as many abortions as there are (postnatal) deaths, which both a) sounds more shocking and b) is more accurate than the original claim. Commented Jul 23, 2013 at 14:46

The global death rate was 7.99 per thousand people per year. The world population is about 7.1 billion people. This means that about 56 million people die a year, excluding abortions. This approximately aligns with the estimate for the number of postnatal deaths in the image.

The Guttmacher Institute estimates that there were 43.8 million abortions in 2008. This data is a little older than the rest, but it's what I could find. assuming the same number of deaths in 2008 as in 2012, That means that abortions comprise about 43% of all deaths (if we define death to include abortions).

If we define deaths to include abortions (but to not include miscarriages), this image is approximately accurate as far as I can tell.

  • 34
    But if we're counting abortions as deaths shouldn't we also be counting miscarriages and conceptions that don't go to term? Only a third of fertilised eggs go to term so that's around 260 million "deaths" a year. Even if we go from implanted eggs, that's another 65 million. Commented Jul 23, 2013 at 6:53
  • 33
    The numbers in this answer are right. I dislike the last sentence for the reason Jack mentioned: the image is pretty misleading because it arbitrarily counts abortions, but not miscarriages as deaths. Commented Jul 23, 2013 at 8:29
  • 12
    I totally agree, but for the purposes of answering the question, I didn't get to make that call.
    – Publius
    Commented Jul 23, 2013 at 14:39
  • 8
    On the other hand, suppose you define abortion as deliberate killing and only compare to executions, murders and manslaughters per capita. Then you don't have to count miscarriages, which are deaths due to natural causes. The number of murders and executions is lower than the number of deaths, so abortions should be far larger than 43%, so a more sensationalist headline can be written, like "Did you know abortions constitute 95% of deliberate human killings worldwide?"
    – Kaz
    Commented Jul 23, 2013 at 21:07
  • 5
    So in other words, it's accurate as long as you use highly selective criteria for what does and doesn't count as a "death"? Commented Jul 24, 2013 at 8:47

The numbers are correct. The definitions, and thus the argument, are not.

As in both Avi's and Konrad's answers, the number of postnatal deaths and the number of abortions in this poster are correct, given the latest information on world population, mortality rate, and number of abortions. However, they were arrived at by ignoring the fact that the mortality rate, which is the number per hundred of deaths as defined by the WHO, does not include abortions, because an abortion is not a "death". It also, as both answers have said, does not include miscarriages, or the grey area of stillbirths (the definition varies by jurisdiction, with some measurements including them as miscarriages, others including them as postnatal mortalities, and still others differentiating them entirely).

These four results - live birth, stillbirth, miscarriage or abortion - are the only possibilities for an implanted egg and a viable pregnancy. I cannot find a worldwide stillbirth rate, but the U.S. rate is one in 115 pregnancies[1]. With the U.S. pregnancy rate at an average of 105.5 per 1,000 women (so 10.55% of women are pregnant at any given time)[2], if we extrapolate these out to the world female population of 3.402 billion[3], the number of stillbirths annually is 3,402,096,414 * .1055 * .008696 = 3.12 million. That is a very low estimate considering that both the U.S. pregnancy rate (due to economic factors both positive and negative) and stillbirth rate (due mainly to healthcare) are relatively low compared to the rest of the world. The annual number of miscarriages worldwide, borrowing from Konrad's answer, is 135.97 million.

For a fertilized egg, which for many pro-life advocates is where life begins, there is a fifth option; non-implantation, which is the case for up to one-half of all fertilized human eggs according to some estimates[4]. Given this estimate, the number of non-implantations annually can be as high as the total number of pregnancies itself.

If the U.S. pregnancy rate were extrapolated to the world female population, then over a nine-month period, there would be on the order of 358 million non-implantations of a fertilized egg, which would become nonviable (die) and the woman would continue her monthly cycle. Pushed out to the full year, that's 478.5 million women per year who, in the eyes of those for whom life begins at fertilization, are murderers.

Lastly, the OP's chart ignores medically-necessary abortions (because the position of most pro-life advocates is that there are no such situations, which is a position that is patently false given medical knowledge to date). A large contributor, though not the only one, is ectopic pregnancy (the egg implants in the fallopian tube instead of the uterine wall). That occurs at a rate of 19.7 per 1000 pregnancies[4], which if we extrapolate from the U.S. pregnancy rate of 10.55% of women and the total world female population of 3.4 billion, is 3402096414 * .1055 * .0197 ~= 7.1 million ectopic pregnancies annually worldwide (and likely more as the pregnancy rate I used is U.S.-specific and thought to be lower than the world average; unfortunately I can't find a good figure for overall world pregnancy rates). Ectopic pregnancies are 100% fatal for the embryo with or without treatment, and without medical treatment, a ruptured ectopic pregnancy is 100% fatal for the mother, so when an ectopic pregnancy is detected prior to rupture, the treatment is a drug-induced abortion using methotrexate, and if a rupture has occured, the fetus is removed as a natural consequence of repairing the damage. Either way, the insurance billing codes that document the medical procedures performed (and are the primary source of statistical data on such things) will include a code for a medical abortion.

This is not the only such indication for a "therapeutic" abortion. Toxcemia/eclampsia calls for immediate delivery of the baby as soon as the mother is stable, whether the baby would be expected to survive or not. Jurisdictions differ as to when such a delivery, resulting in the death of the baby, is a live birth and infant mortality (resulting in filing of a birth and death certificate for the baby), or an abortion by preterm induction (in which case neither a birth certificate nor a death certificate would be issued). Miscarriages, placental abruptions, and other complications of pregnancy often do not result in the woman's body naturally expelling the pregnancy; these situations require medical intervention to "evacuate the pregnancy", and are indistinguishable in terms of the procedure from a surgical abortion of a viable pregnancy and so would include nearly identical procedural codes.

  • Can you add a reference for the incidence of ectopic pregnancies? Also, could you do the sums to show which percentage of the total (post-natal deaths, miscarriages, failed oocyte implantation, medically indicated abortions, other abortions) is due to not medically indicated abortions? This would be very helpful. After a rough estimate based on my numbers and failure to implant (but not considering the other factors) I would expect to end up with a number around 5%, but this might be far off. Commented Jul 23, 2013 at 18:35
  • Unfortunately I can't find good numbers for the total number of abortions performed in any country, much less worldwide, that were attested by the physician performing it as necessary for the life and health of the mother. These cases are a gray area that may or may not be called abortions, depending on who's telling whom. terms like "preterm delivery", "fetal evacuation" etc are used in many cases where the fetus is already dead or would not be classed as a live birth even if technically alive at the time of separation from the mother.
    – KeithS
    Commented Jul 23, 2013 at 18:53
  • The 2%, or thereabouts, rate of ectopic pregnancies seems to be a well-accepted number. Here's just one source which puts it at 1.97%: aafp.org/afp/2000/0215/p1080.html
    – KeithS
    Commented Jul 23, 2013 at 18:54
  • A good place to start might be the number of abortions performed in a facility other than a hospital or critical care center; it is logical (though possibly naive) to think that if the mother's life is threatened, she won't be in a Planned Parenthood clinic.
    – KeithS
    Commented Jul 23, 2013 at 19:06
  • The argument is that the relative number of abortions to deaths is many. All the answers here prove that correct.
    – user11643
    Commented Nov 19, 2017 at 3:44

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .