There is a good case that abortion is morally impermissible – or at least there is significant moral uncertainty.
Even if these arguments fail, abortion could still be a matter of serious concern for EAs (e.g. because fetuses could have significant but not full moral status, or because there are ways to reduce abortions without punishing women for getting them). Put another way, even if one believes abortion is permissible, it likely remains a comparable problem to any problem of infant mortality – but with even more lost life-years, and occurring on a much larger scale than infant mortality.
This is a very important problem, given the life lost and the scale of the problem – tens of millions of abortions around the world each year.
Outside the US, the problem is virtually unchallenged – and even in the US, there are high-impact sectors with minimal anti-abortion sentiment or efforts.
The issue is more tractable than one might think, for various reasons: it is a highly neglected area in most of the world; there are effective and popular policy interventions even in the most pro-choice countries, but even more so in more pro-life countries; progress can be made even without policy interventions; even small reductions in the abortion rate save huge numbers of lives; many people are open to changing their minds about abortion and can do so in a relatively short time; etc.
If you are one of the ~15% of religious EAs, you probably have even more reason to be convinced.
Sex education and contraception may or may not work depending on the case.
Introduction
I realise this is a sensitive topic. In the developed world around 1 in 3 women have an abortion in their lifetime, meaning it is likely people reading this post have been, or will be, personally invested and affected by the topic of abortion. Please forgive me if I fail to address it in a sufficiently sensitive way, and know that this was not my intention. There is, of course, so much more to say about this, but I wanted to try and keep the post relatively short.
I wrote most of this up a couple of years ago but never got round to posting it until Ariel Simnegar’s post, which encouraged me to refine it a little and share here. Though I’ve always been on the fringes of EA and certainly don’t consider myself to be up to date on EA thinking, it was EA thinking that originally made me passionate about this area some years ago, subsequently focusing my academic research on it.
I think this is an important topic for effective altruists to wrestle with, for various reasons: a) if I am right, this one of the most important, neglected, and tractable problems facing humans in the near term; b) as Ariel Simnegar previously pointed out, effective altruists have typically been pretty open to considering neglected causes and particularly neglected communities – animals, future people, etc; c) so much popular discourse around abortion is hostile and badly reasoned, and effective altruists with a common interest in improving the world can improve the calibre of conversation a lot; d) effective altruists have also been open to the idea that morality can involve serious sacrifices of one’s own welfare (even the permanent sacrifice of one’s organs, for some EAs).
Inevitably, as a blog post this discussion will have to cut out the large majority of relevant literature (especially on moral considerations), but I have tried to collect a load of the most commonly asked questions (along with my answers) at https://calumsblog.com/abortion-qa/. Please do get in touch if you would like further references/resources on these questions.
The first-order ethics of abortion
Arguably, for moral uncertainty reasons (given what is at stake), we should expect the arguments for abortion’s permissibility to be pretty seriously compelling before dropping this issue as a serious priority.
Arguments for abortion access can be divided roughly into 3 groups: arguments denying the full moral status of the child (Singer, McMahan, Tooley, etc.); arguments that abortion may be morally or at least legally justifiable even if the child has full moral status/a right to life (Thomson’s violinist[1]); and arguments that even if abortion is wrong and unjust, restricting abortion access has worse consequences than legalising it – or, at least, in some way intolerable consequences.
There are at least somewhat convincing (in my view, compelling) responses to all these arguments (i.e., at least enough to generate significant moral risk).[2] Denial of fetal personhood typically leads to implausible conclusions regarding how we may treat infants and severely disabled humans, and arguably to a denial of human equality even among non-disabled adults. Even if these conclusions are accepted, most people would accept that these are appreciable bullets to bite – especially for those effective altruists who are invested in preventing infant mortality.
Arguments that abortion is permissible even if the child has full moral status typically rely on the claims that abortion is letting die, rather than killing, and that there is no duty to assist the child to rescue it from death. Both claims are needed for abortion to be permissible. But at least one – I argue both – are subject to convincing challenges. Pro-choice legal scholar Kate Greasley has argued that abortion is most plausibly killing rather than letting die,[3] and feminist philosopher Gina Schouten has argued that distinctly feminist considerations imply that we have more duties to help others in need than Thomson’s violinist argument might suggest.[4] Effective altruists, I suggested earlier, might be particularly sympathetic to this latter argument. Other responses to Thomson highlight various other disanalogies between pregnancy and the violinist situation:[5]
In most cases of abortion, the woman is responsible for both the child’s neediness and their intimate biological relationship with the woman – unlike the violinist case.
In the case of abortion, the woman is the mother of the child[6] – unlike the violinist case.[7]
The violinist is in an unnatural situation and being hooked up to the stranger is an unnatural position – by contrast, the fetus is exactly where she is supposed to be in her ‘natural habitat’.
Finally, arguments regarding the negative consequences of banning abortion have, I suggest, been considerably exaggerated – or even reversed in some cases. Since most of these consequences are argued to fall on women, I discuss this in the fuller section on the impact on women below.
Now, an important consideration for EAs is that even if some of the arguments for abortion’s permissibility (moral or legal) are sound, they do not necessarily make pro-life/anti-abortion work in general inappropriate or worthless. Whether or not animals have full moral status, they are worthy of our moral attention. Even if women could permissibly kill or refuse to rescue (depending on how one parses the situation) their child while in the womb, reducing the abortion rate in ways which do not infringe on women’s bodily autonomy is valuable (if the child still has some moral value). And likewise, even if banning abortion has overall worse outcomes, pro-life work which does not focus on legal restrictions is still valuable (for example, the types of interventions discussed in Ariel’s post and its comments).
To summarise thus far: for moral uncertainty reasons, the arguments for abortion should be pretty seriously compelling in order for us to reject abortion as a serious priority. I humbly suggest that, on the contrary, there are powerful counterarguments – at least enough to cast reasonable doubt on abortion’s permissibility. Moreover, even if the arguments for abortion were compelling, abortion might still be a serious priority, so long as embryos/fetuses have some significant moral value. Only arguments suggesting that they have very limited or no value would suffice to undermine the seriousness of abortion – and those arguments would, I suggest, need to be seriously compelling.
I take a broadly rights-based approach to this question, and I realise that not everyone shares the same broad ethical framework. Still, these considerations should be convincing to most of those who accept anything like the framework exemplified in the post-WW2 liberal democratic human rights movement.
But those who accept a more consequentialist framework could also accept a similar argument, so long as embryos/fetuses are morally valuable, or have a valuable future ahead of them. Indeed, there might be ways in which certain consequentialist views could enhance this case. For example, for certain relatively popular EA views with pronatalist implications, abortion could be seriously problematic even if no morally considerable individual is harmed. And for consequentialists in general, the force of rights-based (e.g. bodily autonomy) justifications for abortion may be significantly diminished (though not entirely nullified).
How effective is anti-abortion advocacy?
While there is (in my view) a commendable case for opposing abortion (an action I leave intentionally broad/vague), effective altruists are famously (and often justifiably) averse to political controversies because of their intractability. In many cultures, especially the US, abortion is seemingly in such a stalemate that one might doubt whether a significant impact can be made. I want to suggest that there is a very powerful prima facie case for the effectiveness of certain kinds of pro-life work.
Importance
The moral gravity of abortion has already been discussed above. What hasn’t been discussed is the scale of abortion. In the UK, with a population of 67 million, there are over 200,000 abortions every year. In the US, there are 800,000-1,000,000 abortions every year. This amounts to around 1 in 3 women having an abortion during their lifetime (and most abortions are one of multiple for that person). Global figures are more difficult to ascertain; methods for estimating illegal abortion in pro-life countries are generally highly unreliable and significantly inflate the numbers.[8] However, it is reasonable to suppose that there are at least 30,000,000 abortions globally each year (China alone has over 9 million abortions a year, though has a far higher abortion rate than average). But it is a mistake to look only at current abortion numbers without considering possible abortion numbers in future. In much of the world, abortion is still prohibited and stigmatised, which has a large impact on abortion rates (see later). If that were to change, abortion numbers could become far higher still. The relevant question is not only whether the current kinds of abortion happening could be prevented, but also whether likely new kinds of abortion could be prevented.
Even if one has no problem with early abortion, 10% of abortions are after 13 weeks’ gestation, at which point fetuses are fully formed, and likely conscious and able to feel pain.[9] This would amount to at least 3 million relatively late abortions each year globally – and potentially many more if abortion were to become more normalised and permitted around the world.
In many countries, around 20-25% of non-spontaneously aborted pregnancies end in abortion. In some countries, the percentage is much higher still. Hence it is clear that abortion is a major cause of death, if not the leading cause of death. What’s more, abortion ends a life near its beginning, and preventing an abortion is therefore comparable to (or perhaps marginally better than) preventing the death of an infant, ceteris paribus – therefore maximising the life-years saved.
At this point, anyone familiar with Toby Ord’s work will mention the Scourge: that a large proportion of embryos spontaneously abort, so that spontaneous abortion easily swamps the problem of induced abortion. While I agree that spontaneous abortion is a serious problem – which is not entirely unintuitive, given how much misery it causes the many women who suffer it – I do not think we need to conclude that spontaneous abortion is a more serious/resource-worthy cause of death than induced abortion, for a few reasons: 1) the proportion of embryos spontaneously aborting is often significantly exaggerated; 2) spontaneous abortion is not in itself a cause of death – it is just a description of natural death prior to birth, with myriad different causes (and hence less tractable than one might anticipate); 3) given that many spontaneous abortions are caused by serious genetic anomalies, it is not clear how many could be prevented without potentially changing the identity of the embryo/fetus; 4) it is plausible that a powerful way to draw attention to the value of embryos and fetuses is by anti-abortion advocacy; 5) I argue it is better to prevent a deliberate killing than to prevent a natural death, for a variety of reasons.
Neglectedness
While in the US, pro-life/anti-abortion work has significant funding and attention, it is unlikely the same could be said about anywhere else in the world (excepting, perhaps, parts of Latin America and parts of the Vatican). Even within the US, however, it can be argued that certain influential interventions or spheres of influence are relatively neglected – the academic, international development and tech communities, for example.
Outside the US, serious attempts to reduce the abortion rate or prevent the normalisation or legalisation or abortion are minimal in comparison with other significant global problems such as climate change, HIV/AIDS, other transmissible diseases, poverty, and so on. Yet, as we will see in the next section, this does not make the issue intractable, in part because the populations in many parts of the world are very strongly pro-life.
Tractability
This is where many effective altruists are likely to be doubtful, given the political nature of the topic. Nevertheless, a very powerful argument for tractability can be made.
The scale of abortion is one of the main reasons it is so tractable – not in the sense of eliminating the problem entirely (which is nearly impossible), but in the sense of saving many lives. Consider this: an initiative which permanently caused a 1% reduction in the abortion rate in the UK would save 2,000 whole lives a year – 20,000 over a decade, and 200,000 over a century. This is just in the UK, where cost-effective interventions are rare. Globally, a 1% reduction would save at least 200,000 lives a year.
But percentage reductions of far greater than 1% are perfectly possible, and even more so the prevention of 1% increases. Let’s review some of the evidence for this.
It is known that abortion policy affects sexual behaviour such that restricting abortion leads to considerably more caution with respect to sex and contraception – which leads to significantly fewer abortions, as well as various other benefits (e.g. fewer unintended pregnancies, fewer sexually transmitted diseases).
A systematic review by the Guttmacher Institute found that cutting public funding of abortions leads to an 18-37% reduction in the abortion rate, with the single best study finding a 37% reduction.[10] Even in the UK, where the population is overwhelmingly pro-choice, only half the population supports public funding for abortion. Passing a relatively popular law could therefore save up to 74,000 lives a year – and more could be saved even more easily in bigger countries with more cultural opposition to abortion.
Parental involvement laws are thought to reduce abortion rates among minors by around 15-20%. Although minors do not constitute the majority of people having abortions, they do constitute about 10%. Hence parental involvement laws may reduce abortion rates by 1.5-2% overall directly, and likely have a significant further effect by encouraging more responsible sexual behaviour longer term. Again, even in the very pro-choice UK, a strong majority support parental consent laws.
Informed consent laws (e.g. requiring mandatory ultrasound, provision of information about fetal development, and so on) reduce the abortion rate by 3.7-12%.
Even before the imposition of the most impactful gestational limits on abortion in the US, a leading abortion advocate estimated that 4,000 abortions were averted each year due to gestational limits on abortion – these gestational limits being pretty modest by global standards, around viability (20-24 weeks). The overwhelming majority of people in virtually every country in the world support gestational limits on abortion.
Abortion clinic regulations in Texas, by forcing the closure of nearly half the clinics in Texas, reduced the abortion rate by 13%.
Perhaps most significantly, the evidence has shown that in virtually every country in the world, legalising abortion has led to a massive increase in the number of abortions. The UK, for example, had fewer than 20,000 illegal abortions a year prior to legalisation, according to the best estimates, climbing to 175,000 legal abortions by 1973, just 6 years after legalisation. In Ethiopia, while legal abortions climbed to 300,000 after legalisation in 2005, illegal abortions were estimated to have stayed the same – suggesting an increase of 300,000 abortions as a result of legalisation. Examination of birth rates in many different countries suggests that birth rates fall in response to abortion liberalisation, suggesting that legal abortions (which increase rapidly) do not merely replace illegal abortions.
Likewise, studies of women denied abortions find that only 10-50% of those women go on to have an illegal induced abortion, with the majority continuing the pregnancy to term – suggesting that removing access to abortion entirely prevents the majority of abortions, though of course not all.
Hence even in strongly pro-choice countries, there are a variety of measures which are supported by the general public and which could significantly reduce the abortion rate. Other measures, such as clamping down on coerced abortions, removing caps on benefits above a certain number of children, and improving flexible working for parents (particularly mothers) have not only popular but almost unanimous support among both pro-lifers and pro-choicers.
In pro-life countries the prospects are even stronger: merely by delaying legalisation of abortion for just one year, for example, over 180,000 lives in the UK could have been saved. In some countries, abortion enjoys minimal popular support but yet has already been legalised or could be legalised in the near future, due to lack of investment in anti-abortion efforts. Even simply delaying legalisation by a year could easily save tens or hundreds of thousands of lives. The opportunities for saving enormous numbers of lives are, therefore, particularly strong in culturally pro-life countries.
But even in pro-choice parts of the world, opportunities are still significant to save lives cost-effectively. There are at least three reasons for this, in addition to the sheer magnitude of the problem.
The first reason is that pro-life advocacy can take a wide variety of forms: political advocacy, education, voting, letter-writing, public speaking, pregnancy resource centres, sidewalk counselling, research, earning to give, convincing others, civil service or other specialist work, and so on. So even in countries where some methods of pro-life work seem relatively ineffective (perhaps because, as in the US, pro-life work is not neglected), there are many possible ways to reduce the abortion rate, on a population level and on an individual level. Consider: 1 in 3 women in many Western countries will have an abortion, very often multiple abortions, in their lifetime. It seems likely that even by giving a talk to a potentially sympathetic audience, or having a number of individual conversations about this topic, you could save a significant number of lives.
The second reason is that some people are very open to being convinced about abortion, often being convinced from a single conversation or watching a video – meaning you could make some difference with fairly minimal effort – even in the course of your ordinary social life, in some cases.
The third reason is that people can become very passionate about abortion. This makes a certain kind of individual pro-life advocacy especially effective, namely, convincing individuals in influential positions about the topic,[11] and using platforms to promote pro-life messages to large numbers of people. This has the potential for an extremely large multiplying effect through fairly simple and time-inexpensive individual initiatives.
Hence, even in the absence of legal change, effective change is possible.[12] But legal changes are themselves not as unlikely as we might think.
Offsetting for women
There is one final potential concern effective altruists might have about pro-life work, namely, that limiting abortion access is likely to have a detrimental impact on women which offsets the benefits. There is, of course, an enormous amount of literature on the impact of abortion on women, from both positive and negative perspectives, which is impossible to rehearse in detail here. I offer a slightly more expanded version of my case that abortion has (at the very least) not substantially improved the lives of women on average here. I also have a short paper (‘Defending Dobbs v Jackson: the empirical effect of abortion bans on women’) summarising some of this evidence in the International Family Law Journal and can share a draft on request. I will therefore not attempt the impossible task of a comprehensive review of this topic in this short blog post. But let me briefly point towards some of the empirical evidence alleviating some of these concerns.
Firstly, of course, not all interventions described here involve limiting abortion access. One could still attempt to prevent abortions in ways which do not restrict – and which may indeed enhance – women’s choice, for example, by advocating for laws ensuring that women are provided with financial support and the relevant information regarding it prior to obtaining an abortion.
Secondly, the mental health evidence is increasingly clear that abortion is not associated with mental health benefits compared to continuing an unwanted pregnancy, and the best available evidence suggests that it is associated with poorer mental health outcomes compared to continuing an unwanted pregnancy, after adjusting for confounding factors. See here for a more detailed discussion of the psychiatric evidence.
Thirdly, the argument that women will still get abortions, only unsafe ones, is predicated in part on the idea that restrictions on abortion don’t prevent abortions in the first place. As evidenced above, this is demonstrably false. But clearly some abortions do occur even when abortion is illegal, and some of these are unsafe for the woman.
There is a huge amount to say about the evidence on this question (I am working on 5 or 6 more papers on the topic at the moment), but the basic gist is: 1) the numbers of women dying from induced abortion are frequently long outdated, or in some cases even fabricated; 2) these numbers typically include spontaneous abortion (i.e. miscarriage) and attribute deaths from miscarriage to deaths from ‘unsafe abortion’, inflating the numbers significantly; sometimes ectopic pregnancies are also included; 3) the latest evidence suggests that even in sub-Saharan Africa, often thought to be the region most afflicted by deaths from unsafe abortion, about 5% of maternal deaths are from induced and spontaneous abortion combined; 4) the empirical evidence from particular countries suggests that legalising abortion does not significantly improve the situation (happy to discuss particular examples in detail) and sometimes makes it worse; 5) deaths from abortion are a function of infrastructure, not law: pro-life countries/regions with good healthcare (e.g. Chile, Poland, Malta, South Korea (until recently), Ireland (until recently), North Africa, UAE, and almost all of Europe pre-legalisation) have very few, in many cases zero, deaths from abortion – Malta and Poland have the lowest maternal mortality ratios in the world, for example – while pro-choice countries with relatively poor healthcare (South Africa, India, Ethiopia, Zambia, Rwanda, etc.) still have high numbers of maternal deaths from abortion; 6) as a result of improving post-abortion care and safer illegal abortion (particularly misoprostol), illegal abortion is far safer than it was even a couple of decades ago. For all these reasons, the evidence that restricting abortion worsens maternal mortality is scarce (again, I am happy to discuss this evidence in more detail) and in some cases, restrictions seem to improve the situation.
Finally, the evidence for the financial wellbeing and happiness of women as a result of large-scale abortion access is strikingly slim, given the ostensible benefits. While many feminist writers have written book-length works on this topic, given the limited space I can only point to Akerlof, Yellen and Katz’s seminal work suggesting that the expansion of abortion has led to significant family breakdown, leading in turn to poverty and inequality, with particularly negative economic outcomes for women. Likewise, the declining happiness of women since the 1970s – in absolute terms and relative to men – has been much remarked upon. I say a bit more regarding financial and career outcomes here.
It is impossible to do justice to the interaction between abortion access and women’s wellbeing in this short blog post, but I hope this is enough for now to show that it is unlikely – or at least far from clear – that the benefits of reducing abortion rates will be significantly offset by the impact of that reduction on women.
Conclusion
In summary, there are convincing arguments that abortion is a serious issue for effective altruists, and even if one is persuaded that abortion is morally or legally justifiable, reducing abortion rates should still be a priority area. But if the pro-life arguments are correct, then it is arguably harder to find an issue that is more serious – at least for human beings in the near term – than abortion. But it is partly for this reason that it is so tractable: even tiny reductions in the abortion rate globally and domestically can save large numbers of lives. Much bigger reductions can be made by interventions which are popular even in strongly pro-choice countries, a fortiori in countries with significantly more pro-life sentiment. A prima facie assessment of the cost-effectiveness of pro-life advocacy therefore appears to highly commend it.
Put another way: any EAs who are invested in preventing infant mortality should consider abortion a similarly pressing priority. Such EAs are likely to be unpersuaded by anti-personhood arguments (considering them either false or irrelevant, since they would undermine the case against infant mortality as well) – and as long as fetuses or their future lives have significant moral value – abortion is a serious priority.
Again, I would like to reiterate that I know how delicate and personally invested this subject can be. But I don’t think – as with vegetarianism and other EA areas – the fear of causing offence or appearing like a moraliser should prevent us from discussing potentially extremely important moral issues. But I hope (and will try my best) that we can discuss it with humility and respect, and I am grateful the EA community has so far given me every reason to expect that again here. Thanks—and thank you to Ariel Simnegar for his helpful comments on a draft of this piece.
Appendix 1: Religion and abortion
It is worth noting – since a non-trivial (~15%) number of effective altruists have religious commitments – that the world’s largest religious traditions have generally been opposed to abortion, which would add extra weight to this case for those with those religious commitments. The Christian tradition has been, from the 1st century until the 20th, unanimously opposed to abortion, except when the mother’s life is at risk. Likewise, Judaism historically opposed abortion except when the mother’s life is at risk.[13] There is a broad consensus among Islamic scholars that abortion is impermissible except in rare cases, even prior to ensoulment.[14] While there has been some dissent from these positions, this has been limited, especially in Christianity and Islam. It is true that in each of the Abrahamic faiths, there has been some historical debate about when ensoulment occurs, and therefore whether abortion is specifically murder. But they have traditionally maintained that even if it is not murder, it is seriously wrong.[15]
Hence those who belong to a broadly orthodox version of these religions will have extra reason to consider abortion a serious problem and a priority area.
Appendix 2: Contraception and sex education
One obvious response to all this is to say that we should reduce abortions, and a great way to do this is by promoting contraception and sex education. The evidence base on these is too large to describe in great detail, but I want to at least offer some very brief cautions about this approach (and happy to discuss in more detail in the comments).
Both interventions have mixed effects: they can, in theory, prevent pregnancies and therefore abortions (through obvious mechanisms), but they can also increase both on a population level (e.g. by incentivising risky sex and thus causing more unplanned pregnancies – see the literature on risk compensation). Hence their effect must be evaluated on the basis of empirical evidence, rather than just in theory.
The effect of contraception appears to be varied. When promoted to individuals who are already high-risk, it is probably effective to reduce pregnancies and abortions. But when promoted in a population more generally, which may be low-risk, it can incentivise risky sex and make the problem worse. Hence in some countries, abortion declines as contraceptive prevalence rises (typically former Soviet bloc countries which had already liberalised on sex/birth control and used abortion as birth control before hormonal contraception was available), and in other countries, they have grown together.
Systematic reviews on emergency contraception access have repeatedly found that it has no impact on the pregnancy/abortion rate, and a Cochrane review of randomised trials found no impact of contraception-promoting interventions on the pregnancy rate.
Finally, contraception access is already surprisingly good: unmet need for contraception is very low even in the developing world, and only a very small percentage of this is because of lack access. There is much more to say, but these are just some cautions regarding contraception promotion as a solution.
Likewise, the impact of sex education depends on the kind of sex education. We can expect that some kinds of sex education might reduce unplanned pregnancies, while other kinds might increase them, or have no effect. A Cochrane review of randomised trials found no overall effect of sex education on pregnancy rates – the trial which did show a huge reduction was in Cabezón (2005) using TeenStar – a more conservative type of sex education. Another Cochrane review found similar results – e.g. Clark (2005) did find that an educational intervention worked in delaying initiation of sex, but this did not meet statistical significance, and had nothing to do with sex education specifically. A more recent study has found that sex education mandates were associated with increased teen pregnancies – and that this result was mitigated by parental opt-out policies. Of course, there is much more to be said here (including about individual studies in the reviews mentioned), and I am happy to discuss the details in the comments. But for these sorts of reasons I am not as optimistic about (at least standard) sex education as a remedy.
Footnotes
[1] Though Thomson doesn’t quite go so far as to say that abortion is morally permissible – only that it does not violate the child’s right to life. Thomson allows that abortion could be indecent, callous, and otherwise immoral, but argues that it is not so qua a violation of the right to life.
[2] Perhaps of particular interest to effective altruists is Don Marquis’ seminal article, which argues that killing is wrong primarily because it deprives an individual of a future life of value (or ‘future like ours’). Since most abortions involve depriving an individual of a future like ours, abortion is wrong for the same reason that killing humans in general is wrong. But if Marquis’ argument is correct, abortion may be worse, since it involves depriving someone of more life than almost any other form of killing. From an effective altruist perspective, by saving a life from abortion, one saves almost the entire life. My view is that this is not the primary reason killing is wrong, but it is an important one (and an important reason that death is bad).
[3] Of course, many consequentialists are sympathetic to rejecting this distinction altogether, which would mean any justification for abortion would (in effect) have to be a sufficient justification for killing – a relatively high bar.
[4] I have a paper arguing broadly the same conclusion, but with respect to other progressive emphases, such as wealth redistribution and the rectification of social injustices.
[6] I use terminology like ‘child’ here because these arguments for abortion are supposed to be effective even if one concedes the full moral status of the fetus.
[7] Of course, this would not apply in the case of rape, but these are such a tiny (though important) percentage of abortions that it would not significantly affect the urgency of preventing abortion more generally, assuming there is such an urgency.
[10] It might be thought that the Guttmacher Institute (pro-choice) has a bias towards inflating the harm of abortion bans – but I suggest the opposite is true, since their public messaging tends to claim that abortion restrictions are totally ineffective (and therefore should be repealed) – although of course this is inconsistent with their research.
[11] For example, consider the impact Singer’s writing had on influential animal advocate Brian Tomasik.
[13] Certainly prior to the Talmud. The Talmudic teaching on abortion and related topics is somewhat complicated, but generally supports a pro-life perspective (e.g. including fetuses in the prohibition on killing in Genesis 9:6).
[14] Which is held to occur at 40, 80, or 120 days by different scholars.
[15] Again, I am happy to provide more background evidence for these claims on Christianity and historically in Judaism – as well as further comments on the scriptural support underpinning these traditional positions.
The effective altruist case for pro-life/anti-abortion advocacy
Summary
There is a good case that abortion is morally impermissible – or at least there is significant moral uncertainty.
Even if these arguments fail, abortion could still be a matter of serious concern for EAs (e.g. because fetuses could have significant but not full moral status, or because there are ways to reduce abortions without punishing women for getting them). Put another way, even if one believes abortion is permissible, it likely remains a comparable problem to any problem of infant mortality – but with even more lost life-years, and occurring on a much larger scale than infant mortality.
This is a very important problem, given the life lost and the scale of the problem – tens of millions of abortions around the world each year.
Outside the US, the problem is virtually unchallenged – and even in the US, there are high-impact sectors with minimal anti-abortion sentiment or efforts.
The issue is more tractable than one might think, for various reasons: it is a highly neglected area in most of the world; there are effective and popular policy interventions even in the most pro-choice countries, but even more so in more pro-life countries; progress can be made even without policy interventions; even small reductions in the abortion rate save huge numbers of lives; many people are open to changing their minds about abortion and can do so in a relatively short time; etc.
If you are one of the ~15% of religious EAs, you probably have even more reason to be convinced.
Sex education and contraception may or may not work depending on the case.
Introduction
I realise this is a sensitive topic. In the developed world around 1 in 3 women have an abortion in their lifetime, meaning it is likely people reading this post have been, or will be, personally invested and affected by the topic of abortion. Please forgive me if I fail to address it in a sufficiently sensitive way, and know that this was not my intention. There is, of course, so much more to say about this, but I wanted to try and keep the post relatively short.
I wrote most of this up a couple of years ago but never got round to posting it until Ariel Simnegar’s post, which encouraged me to refine it a little and share here. Though I’ve always been on the fringes of EA and certainly don’t consider myself to be up to date on EA thinking, it was EA thinking that originally made me passionate about this area some years ago, subsequently focusing my academic research on it.
I think this is an important topic for effective altruists to wrestle with, for various reasons: a) if I am right, this one of the most important, neglected, and tractable problems facing humans in the near term; b) as Ariel Simnegar previously pointed out, effective altruists have typically been pretty open to considering neglected causes and particularly neglected communities – animals, future people, etc; c) so much popular discourse around abortion is hostile and badly reasoned, and effective altruists with a common interest in improving the world can improve the calibre of conversation a lot; d) effective altruists have also been open to the idea that morality can involve serious sacrifices of one’s own welfare (even the permanent sacrifice of one’s organs, for some EAs).
Inevitably, as a blog post this discussion will have to cut out the large majority of relevant literature (especially on moral considerations), but I have tried to collect a load of the most commonly asked questions (along with my answers) at https://calumsblog.com/abortion-qa/. Please do get in touch if you would like further references/resources on these questions.
The first-order ethics of abortion
Arguably, for moral uncertainty reasons (given what is at stake), we should expect the arguments for abortion’s permissibility to be pretty seriously compelling before dropping this issue as a serious priority.
Arguments for abortion access can be divided roughly into 3 groups: arguments denying the full moral status of the child (Singer, McMahan, Tooley, etc.); arguments that abortion may be morally or at least legally justifiable even if the child has full moral status/a right to life (Thomson’s violinist[1]); and arguments that even if abortion is wrong and unjust, restricting abortion access has worse consequences than legalising it – or, at least, in some way intolerable consequences.
There are at least somewhat convincing (in my view, compelling) responses to all these arguments (i.e., at least enough to generate significant moral risk).[2] Denial of fetal personhood typically leads to implausible conclusions regarding how we may treat infants and severely disabled humans, and arguably to a denial of human equality even among non-disabled adults. Even if these conclusions are accepted, most people would accept that these are appreciable bullets to bite – especially for those effective altruists who are invested in preventing infant mortality.
Arguments that abortion is permissible even if the child has full moral status typically rely on the claims that abortion is letting die, rather than killing, and that there is no duty to assist the child to rescue it from death. Both claims are needed for abortion to be permissible. But at least one – I argue both – are subject to convincing challenges. Pro-choice legal scholar Kate Greasley has argued that abortion is most plausibly killing rather than letting die,[3] and feminist philosopher Gina Schouten has argued that distinctly feminist considerations imply that we have more duties to help others in need than Thomson’s violinist argument might suggest.[4] Effective altruists, I suggested earlier, might be particularly sympathetic to this latter argument. Other responses to Thomson highlight various other disanalogies between pregnancy and the violinist situation:[5]
In most cases of abortion, the woman is responsible for both the child’s neediness and their intimate biological relationship with the woman – unlike the violinist case.
In the case of abortion, the woman is the mother of the child[6] – unlike the violinist case.[7]
The violinist is in an unnatural situation and being hooked up to the stranger is an unnatural position – by contrast, the fetus is exactly where she is supposed to be in her ‘natural habitat’.
Finally, arguments regarding the negative consequences of banning abortion have, I suggest, been considerably exaggerated – or even reversed in some cases. Since most of these consequences are argued to fall on women, I discuss this in the fuller section on the impact on women below.
Now, an important consideration for EAs is that even if some of the arguments for abortion’s permissibility (moral or legal) are sound, they do not necessarily make pro-life/anti-abortion work in general inappropriate or worthless. Whether or not animals have full moral status, they are worthy of our moral attention. Even if women could permissibly kill or refuse to rescue (depending on how one parses the situation) their child while in the womb, reducing the abortion rate in ways which do not infringe on women’s bodily autonomy is valuable (if the child still has some moral value). And likewise, even if banning abortion has overall worse outcomes, pro-life work which does not focus on legal restrictions is still valuable (for example, the types of interventions discussed in Ariel’s post and its comments).
To summarise thus far: for moral uncertainty reasons, the arguments for abortion should be pretty seriously compelling in order for us to reject abortion as a serious priority. I humbly suggest that, on the contrary, there are powerful counterarguments – at least enough to cast reasonable doubt on abortion’s permissibility. Moreover, even if the arguments for abortion were compelling, abortion might still be a serious priority, so long as embryos/fetuses have some significant moral value. Only arguments suggesting that they have very limited or no value would suffice to undermine the seriousness of abortion – and those arguments would, I suggest, need to be seriously compelling.
I take a broadly rights-based approach to this question, and I realise that not everyone shares the same broad ethical framework. Still, these considerations should be convincing to most of those who accept anything like the framework exemplified in the post-WW2 liberal democratic human rights movement.
But those who accept a more consequentialist framework could also accept a similar argument, so long as embryos/fetuses are morally valuable, or have a valuable future ahead of them. Indeed, there might be ways in which certain consequentialist views could enhance this case. For example, for certain relatively popular EA views with pronatalist implications, abortion could be seriously problematic even if no morally considerable individual is harmed. And for consequentialists in general, the force of rights-based (e.g. bodily autonomy) justifications for abortion may be significantly diminished (though not entirely nullified).
How effective is anti-abortion advocacy?
While there is (in my view) a commendable case for opposing abortion (an action I leave intentionally broad/vague), effective altruists are famously (and often justifiably) averse to political controversies because of their intractability. In many cultures, especially the US, abortion is seemingly in such a stalemate that one might doubt whether a significant impact can be made. I want to suggest that there is a very powerful prima facie case for the effectiveness of certain kinds of pro-life work.
Importance
The moral gravity of abortion has already been discussed above. What hasn’t been discussed is the scale of abortion. In the UK, with a population of 67 million, there are over 200,000 abortions every year. In the US, there are 800,000-1,000,000 abortions every year. This amounts to around 1 in 3 women having an abortion during their lifetime (and most abortions are one of multiple for that person). Global figures are more difficult to ascertain; methods for estimating illegal abortion in pro-life countries are generally highly unreliable and significantly inflate the numbers.[8] However, it is reasonable to suppose that there are at least 30,000,000 abortions globally each year (China alone has over 9 million abortions a year, though has a far higher abortion rate than average). But it is a mistake to look only at current abortion numbers without considering possible abortion numbers in future. In much of the world, abortion is still prohibited and stigmatised, which has a large impact on abortion rates (see later). If that were to change, abortion numbers could become far higher still. The relevant question is not only whether the current kinds of abortion happening could be prevented, but also whether likely new kinds of abortion could be prevented.
Even if one has no problem with early abortion, 10% of abortions are after 13 weeks’ gestation, at which point fetuses are fully formed, and likely conscious and able to feel pain.[9] This would amount to at least 3 million relatively late abortions each year globally – and potentially many more if abortion were to become more normalised and permitted around the world.
In many countries, around 20-25% of non-spontaneously aborted pregnancies end in abortion. In some countries, the percentage is much higher still. Hence it is clear that abortion is a major cause of death, if not the leading cause of death. What’s more, abortion ends a life near its beginning, and preventing an abortion is therefore comparable to (or perhaps marginally better than) preventing the death of an infant, ceteris paribus – therefore maximising the life-years saved.
At this point, anyone familiar with Toby Ord’s work will mention the Scourge: that a large proportion of embryos spontaneously abort, so that spontaneous abortion easily swamps the problem of induced abortion. While I agree that spontaneous abortion is a serious problem – which is not entirely unintuitive, given how much misery it causes the many women who suffer it – I do not think we need to conclude that spontaneous abortion is a more serious/resource-worthy cause of death than induced abortion, for a few reasons: 1) the proportion of embryos spontaneously aborting is often significantly exaggerated; 2) spontaneous abortion is not in itself a cause of death – it is just a description of natural death prior to birth, with myriad different causes (and hence less tractable than one might anticipate); 3) given that many spontaneous abortions are caused by serious genetic anomalies, it is not clear how many could be prevented without potentially changing the identity of the embryo/fetus; 4) it is plausible that a powerful way to draw attention to the value of embryos and fetuses is by anti-abortion advocacy; 5) I argue it is better to prevent a deliberate killing than to prevent a natural death, for a variety of reasons.
Neglectedness
While in the US, pro-life/anti-abortion work has significant funding and attention, it is unlikely the same could be said about anywhere else in the world (excepting, perhaps, parts of Latin America and parts of the Vatican). Even within the US, however, it can be argued that certain influential interventions or spheres of influence are relatively neglected – the academic, international development and tech communities, for example.
Outside the US, serious attempts to reduce the abortion rate or prevent the normalisation or legalisation or abortion are minimal in comparison with other significant global problems such as climate change, HIV/AIDS, other transmissible diseases, poverty, and so on. Yet, as we will see in the next section, this does not make the issue intractable, in part because the populations in many parts of the world are very strongly pro-life.
Tractability
This is where many effective altruists are likely to be doubtful, given the political nature of the topic. Nevertheless, a very powerful argument for tractability can be made.
The scale of abortion is one of the main reasons it is so tractable – not in the sense of eliminating the problem entirely (which is nearly impossible), but in the sense of saving many lives. Consider this: an initiative which permanently caused a 1% reduction in the abortion rate in the UK would save 2,000 whole lives a year – 20,000 over a decade, and 200,000 over a century. This is just in the UK, where cost-effective interventions are rare. Globally, a 1% reduction would save at least 200,000 lives a year.
But percentage reductions of far greater than 1% are perfectly possible, and even more so the prevention of 1% increases. Let’s review some of the evidence for this.
It is known that abortion policy affects sexual behaviour such that restricting abortion leads to considerably more caution with respect to sex and contraception – which leads to significantly fewer abortions, as well as various other benefits (e.g. fewer unintended pregnancies, fewer sexually transmitted diseases).
A systematic review by the Guttmacher Institute found that cutting public funding of abortions leads to an 18-37% reduction in the abortion rate, with the single best study finding a 37% reduction.[10] Even in the UK, where the population is overwhelmingly pro-choice, only half the population supports public funding for abortion. Passing a relatively popular law could therefore save up to 74,000 lives a year – and more could be saved even more easily in bigger countries with more cultural opposition to abortion.
Parental involvement laws are thought to reduce abortion rates among minors by around 15-20%. Although minors do not constitute the majority of people having abortions, they do constitute about 10%. Hence parental involvement laws may reduce abortion rates by 1.5-2% overall directly, and likely have a significant further effect by encouraging more responsible sexual behaviour longer term. Again, even in the very pro-choice UK, a strong majority support parental consent laws.
Informed consent laws (e.g. requiring mandatory ultrasound, provision of information about fetal development, and so on) reduce the abortion rate by 3.7-12%.
Even before the imposition of the most impactful gestational limits on abortion in the US, a leading abortion advocate estimated that 4,000 abortions were averted each year due to gestational limits on abortion – these gestational limits being pretty modest by global standards, around viability (20-24 weeks). The overwhelming majority of people in virtually every country in the world support gestational limits on abortion.
Abortion clinic regulations in Texas, by forcing the closure of nearly half the clinics in Texas, reduced the abortion rate by 13%.
Perhaps most significantly, the evidence has shown that in virtually every country in the world, legalising abortion has led to a massive increase in the number of abortions. The UK, for example, had fewer than 20,000 illegal abortions a year prior to legalisation, according to the best estimates, climbing to 175,000 legal abortions by 1973, just 6 years after legalisation. In Ethiopia, while legal abortions climbed to 300,000 after legalisation in 2005, illegal abortions were estimated to have stayed the same – suggesting an increase of 300,000 abortions as a result of legalisation. Examination of birth rates in many different countries suggests that birth rates fall in response to abortion liberalisation, suggesting that legal abortions (which increase rapidly) do not merely replace illegal abortions.
Likewise, studies of women denied abortions find that only 10-50% of those women go on to have an illegal induced abortion, with the majority continuing the pregnancy to term – suggesting that removing access to abortion entirely prevents the majority of abortions, though of course not all.
Hence even in strongly pro-choice countries, there are a variety of measures which are supported by the general public and which could significantly reduce the abortion rate. Other measures, such as clamping down on coerced abortions, removing caps on benefits above a certain number of children, and improving flexible working for parents (particularly mothers) have not only popular but almost unanimous support among both pro-lifers and pro-choicers.
In pro-life countries the prospects are even stronger: merely by delaying legalisation of abortion for just one year, for example, over 180,000 lives in the UK could have been saved. In some countries, abortion enjoys minimal popular support but yet has already been legalised or could be legalised in the near future, due to lack of investment in anti-abortion efforts. Even simply delaying legalisation by a year could easily save tens or hundreds of thousands of lives. The opportunities for saving enormous numbers of lives are, therefore, particularly strong in culturally pro-life countries.
But even in pro-choice parts of the world, opportunities are still significant to save lives cost-effectively. There are at least three reasons for this, in addition to the sheer magnitude of the problem.
The first reason is that pro-life advocacy can take a wide variety of forms: political advocacy, education, voting, letter-writing, public speaking, pregnancy resource centres, sidewalk counselling, research, earning to give, convincing others, civil service or other specialist work, and so on. So even in countries where some methods of pro-life work seem relatively ineffective (perhaps because, as in the US, pro-life work is not neglected), there are many possible ways to reduce the abortion rate, on a population level and on an individual level. Consider: 1 in 3 women in many Western countries will have an abortion, very often multiple abortions, in their lifetime. It seems likely that even by giving a talk to a potentially sympathetic audience, or having a number of individual conversations about this topic, you could save a significant number of lives.
The second reason is that some people are very open to being convinced about abortion, often being convinced from a single conversation or watching a video – meaning you could make some difference with fairly minimal effort – even in the course of your ordinary social life, in some cases.
The third reason is that people can become very passionate about abortion. This makes a certain kind of individual pro-life advocacy especially effective, namely, convincing individuals in influential positions about the topic,[11] and using platforms to promote pro-life messages to large numbers of people. This has the potential for an extremely large multiplying effect through fairly simple and time-inexpensive individual initiatives.
Hence, even in the absence of legal change, effective change is possible.[12] But legal changes are themselves not as unlikely as we might think.
Offsetting for women
There is one final potential concern effective altruists might have about pro-life work, namely, that limiting abortion access is likely to have a detrimental impact on women which offsets the benefits. There is, of course, an enormous amount of literature on the impact of abortion on women, from both positive and negative perspectives, which is impossible to rehearse in detail here. I offer a slightly more expanded version of my case that abortion has (at the very least) not substantially improved the lives of women on average here. I also have a short paper (‘Defending Dobbs v Jackson: the empirical effect of abortion bans on women’) summarising some of this evidence in the International Family Law Journal and can share a draft on request. I will therefore not attempt the impossible task of a comprehensive review of this topic in this short blog post. But let me briefly point towards some of the empirical evidence alleviating some of these concerns.
Firstly, of course, not all interventions described here involve limiting abortion access. One could still attempt to prevent abortions in ways which do not restrict – and which may indeed enhance – women’s choice, for example, by advocating for laws ensuring that women are provided with financial support and the relevant information regarding it prior to obtaining an abortion.
Secondly, the mental health evidence is increasingly clear that abortion is not associated with mental health benefits compared to continuing an unwanted pregnancy, and the best available evidence suggests that it is associated with poorer mental health outcomes compared to continuing an unwanted pregnancy, after adjusting for confounding factors. See here for a more detailed discussion of the psychiatric evidence.
Thirdly, the argument that women will still get abortions, only unsafe ones, is predicated in part on the idea that restrictions on abortion don’t prevent abortions in the first place. As evidenced above, this is demonstrably false. But clearly some abortions do occur even when abortion is illegal, and some of these are unsafe for the woman.
There is a huge amount to say about the evidence on this question (I am working on 5 or 6 more papers on the topic at the moment), but the basic gist is: 1) the numbers of women dying from induced abortion are frequently long outdated, or in some cases even fabricated; 2) these numbers typically include spontaneous abortion (i.e. miscarriage) and attribute deaths from miscarriage to deaths from ‘unsafe abortion’, inflating the numbers significantly; sometimes ectopic pregnancies are also included; 3) the latest evidence suggests that even in sub-Saharan Africa, often thought to be the region most afflicted by deaths from unsafe abortion, about 5% of maternal deaths are from induced and spontaneous abortion combined; 4) the empirical evidence from particular countries suggests that legalising abortion does not significantly improve the situation (happy to discuss particular examples in detail) and sometimes makes it worse; 5) deaths from abortion are a function of infrastructure, not law: pro-life countries/regions with good healthcare (e.g. Chile, Poland, Malta, South Korea (until recently), Ireland (until recently), North Africa, UAE, and almost all of Europe pre-legalisation) have very few, in many cases zero, deaths from abortion – Malta and Poland have the lowest maternal mortality ratios in the world, for example – while pro-choice countries with relatively poor healthcare (South Africa, India, Ethiopia, Zambia, Rwanda, etc.) still have high numbers of maternal deaths from abortion; 6) as a result of improving post-abortion care and safer illegal abortion (particularly misoprostol), illegal abortion is far safer than it was even a couple of decades ago. For all these reasons, the evidence that restricting abortion worsens maternal mortality is scarce (again, I am happy to discuss this evidence in more detail) and in some cases, restrictions seem to improve the situation.
Finally, the evidence for the financial wellbeing and happiness of women as a result of large-scale abortion access is strikingly slim, given the ostensible benefits. While many feminist writers have written book-length works on this topic, given the limited space I can only point to Akerlof, Yellen and Katz’s seminal work suggesting that the expansion of abortion has led to significant family breakdown, leading in turn to poverty and inequality, with particularly negative economic outcomes for women. Likewise, the declining happiness of women since the 1970s – in absolute terms and relative to men – has been much remarked upon. I say a bit more regarding financial and career outcomes here.
It is impossible to do justice to the interaction between abortion access and women’s wellbeing in this short blog post, but I hope this is enough for now to show that it is unlikely – or at least far from clear – that the benefits of reducing abortion rates will be significantly offset by the impact of that reduction on women.
Conclusion
In summary, there are convincing arguments that abortion is a serious issue for effective altruists, and even if one is persuaded that abortion is morally or legally justifiable, reducing abortion rates should still be a priority area. But if the pro-life arguments are correct, then it is arguably harder to find an issue that is more serious – at least for human beings in the near term – than abortion. But it is partly for this reason that it is so tractable: even tiny reductions in the abortion rate globally and domestically can save large numbers of lives. Much bigger reductions can be made by interventions which are popular even in strongly pro-choice countries, a fortiori in countries with significantly more pro-life sentiment. A prima facie assessment of the cost-effectiveness of pro-life advocacy therefore appears to highly commend it.
Put another way: any EAs who are invested in preventing infant mortality should consider abortion a similarly pressing priority. Such EAs are likely to be unpersuaded by anti-personhood arguments (considering them either false or irrelevant, since they would undermine the case against infant mortality as well) – and as long as fetuses or their future lives have significant moral value – abortion is a serious priority.
Again, I would like to reiterate that I know how delicate and personally invested this subject can be. But I don’t think – as with vegetarianism and other EA areas – the fear of causing offence or appearing like a moraliser should prevent us from discussing potentially extremely important moral issues. But I hope (and will try my best) that we can discuss it with humility and respect, and I am grateful the EA community has so far given me every reason to expect that again here. Thanks—and thank you to Ariel Simnegar for his helpful comments on a draft of this piece.
Appendix 1: Religion and abortion
It is worth noting – since a non-trivial (~15%) number of effective altruists have religious commitments – that the world’s largest religious traditions have generally been opposed to abortion, which would add extra weight to this case for those with those religious commitments. The Christian tradition has been, from the 1st century until the 20th, unanimously opposed to abortion, except when the mother’s life is at risk. Likewise, Judaism historically opposed abortion except when the mother’s life is at risk.[13] There is a broad consensus among Islamic scholars that abortion is impermissible except in rare cases, even prior to ensoulment.[14] While there has been some dissent from these positions, this has been limited, especially in Christianity and Islam. It is true that in each of the Abrahamic faiths, there has been some historical debate about when ensoulment occurs, and therefore whether abortion is specifically murder. But they have traditionally maintained that even if it is not murder, it is seriously wrong.[15]
I am less familiar with Buddhism and Hinduism, though my understanding is that abortion is generally considered to be a violation of the first precept of Buddhist ethics, and a violation of the ahimsa (non-violence) principle of Hindu ethics.
Hence those who belong to a broadly orthodox version of these religions will have extra reason to consider abortion a serious problem and a priority area.
Appendix 2: Contraception and sex education
One obvious response to all this is to say that we should reduce abortions, and a great way to do this is by promoting contraception and sex education. The evidence base on these is too large to describe in great detail, but I want to at least offer some very brief cautions about this approach (and happy to discuss in more detail in the comments).
Both interventions have mixed effects: they can, in theory, prevent pregnancies and therefore abortions (through obvious mechanisms), but they can also increase both on a population level (e.g. by incentivising risky sex and thus causing more unplanned pregnancies – see the literature on risk compensation). Hence their effect must be evaluated on the basis of empirical evidence, rather than just in theory.
The effect of contraception appears to be varied. When promoted to individuals who are already high-risk, it is probably effective to reduce pregnancies and abortions. But when promoted in a population more generally, which may be low-risk, it can incentivise risky sex and make the problem worse. Hence in some countries, abortion declines as contraceptive prevalence rises (typically former Soviet bloc countries which had already liberalised on sex/birth control and used abortion as birth control before hormonal contraception was available), and in other countries, they have grown together.
Systematic reviews on emergency contraception access have repeatedly found that it has no impact on the pregnancy/abortion rate, and a Cochrane review of randomised trials found no impact of contraception-promoting interventions on the pregnancy rate.
Finally, contraception access is already surprisingly good: unmet need for contraception is very low even in the developing world, and only a very small percentage of this is because of lack access. There is much more to say, but these are just some cautions regarding contraception promotion as a solution.
Likewise, the impact of sex education depends on the kind of sex education. We can expect that some kinds of sex education might reduce unplanned pregnancies, while other kinds might increase them, or have no effect. A Cochrane review of randomised trials found no overall effect of sex education on pregnancy rates – the trial which did show a huge reduction was in Cabezón (2005) using TeenStar – a more conservative type of sex education. Another Cochrane review found similar results – e.g. Clark (2005) did find that an educational intervention worked in delaying initiation of sex, but this did not meet statistical significance, and had nothing to do with sex education specifically. A more recent study has found that sex education mandates were associated with increased teen pregnancies – and that this result was mitigated by parental opt-out policies. Of course, there is much more to be said here (including about individual studies in the reviews mentioned), and I am happy to discuss the details in the comments. But for these sorts of reasons I am not as optimistic about (at least standard) sex education as a remedy.
Footnotes
[1] Though Thomson doesn’t quite go so far as to say that abortion is morally permissible – only that it does not violate the child’s right to life. Thomson allows that abortion could be indecent, callous, and otherwise immoral, but argues that it is not so qua a violation of the right to life.
[2] Perhaps of particular interest to effective altruists is Don Marquis’ seminal article, which argues that killing is wrong primarily because it deprives an individual of a future life of value (or ‘future like ours’). Since most abortions involve depriving an individual of a future like ours, abortion is wrong for the same reason that killing humans in general is wrong. But if Marquis’ argument is correct, abortion may be worse, since it involves depriving someone of more life than almost any other form of killing. From an effective altruist perspective, by saving a life from abortion, one saves almost the entire life. My view is that this is not the primary reason killing is wrong, but it is an important one (and an important reason that death is bad).
[3] Of course, many consequentialists are sympathetic to rejecting this distinction altogether, which would mean any justification for abortion would (in effect) have to be a sufficient justification for killing – a relatively high bar.
[4] I have a paper arguing broadly the same conclusion, but with respect to other progressive emphases, such as wealth redistribution and the rectification of social injustices.
[5] Also worth noting is McMahan’s lengthy treatment of Thomson’s argument in his 2002 book.
[6] I use terminology like ‘child’ here because these arguments for abortion are supposed to be effective even if one concedes the full moral status of the fetus.
[7] Of course, this would not apply in the case of rape, but these are such a tiny (though important) percentage of abortions that it would not significantly affect the urgency of preventing abortion more generally, assuming there is such an urgency.
[8] Details available on request.
[9] Some suggest this is possible earlier (see also recent research on thalamocortical precursors).
[10] It might be thought that the Guttmacher Institute (pro-choice) has a bias towards inflating the harm of abortion bans – but I suggest the opposite is true, since their public messaging tends to claim that abortion restrictions are totally ineffective (and therefore should be repealed) – although of course this is inconsistent with their research.
[11] For example, consider the impact Singer’s writing had on influential animal advocate Brian Tomasik.
[12] See a parallel in vegan advocacy here.
[13] Certainly prior to the Talmud. The Talmudic teaching on abortion and related topics is somewhat complicated, but generally supports a pro-life perspective (e.g. including fetuses in the prohibition on killing in Genesis 9:6).
[14] Which is held to occur at 40, 80, or 120 days by different scholars.
[15] Again, I am happy to provide more background evidence for these claims on Christianity and historically in Judaism – as well as further comments on the scriptural support underpinning these traditional positions.