Is Abortion the Only Issue?

2.1. The Reasoning

A common pro-life argument for prioritizing abortion involves appealing to the large number of abortions. I’m calling this numbers-focused approach “Body Count Reasoning.” It’s seldom or never presented as an explicit, formal argument, so presenting it as one will require some reconstruction. I’ll understand the basic argument as looking like this:

BCR-General:

Some notes. As I’m imagining it, acceptance of premise (1) will be grounded in acceptance of at least three claims. The first is that fetuses are persons. The second is that abortion is (at least pretty frequently) unjust killing. (This second part is important partly because Body Count Reasoning sometimes specifically appeals, not just to the claim that abortion involves lot of deaths, but specifically to the claim that it involves lots of unjust killing, which might be thought worse. The Body Count Reasoner might also appeal to other claims, such as that abortion is legal killing, or killing done by the child’s own mother; I’ll address these moves in Section 4.) Both of these are commitments of the pro-life view.

The third claim behind premise (1) is that the number of abortions is much larger than the number of people killed by other issues. This is empirically uncontroversial (with one caveat: see Section 2.3). One estimate puts the number of abortions worldwide at over fifty-six million annually (Sedgh et al. 2016), a number about equal to the total number of deaths of already-born people from all causes (WHO 2018). In the United States, the Guttmacher Institute estimates that there were around 862,000 abortions in 2017 (Nash & Dreweke 2019). This dwarfs the number of deaths from other issues that receive substantial political attention. For instance, in the US there are around 45,000 firearm deaths (CDC 2017) and 70,000 deaths by drug overdose (Hedegaard et al. 2018) annually. The total number of “deaths of despair”—deaths from suicide and drug and alcohol abuse—is 158,000 annually in the US; the number has increased by around 90,000 from the mid-1990’s (Karma 2020), with the increase driven by socio-economic factors (Case & Deaton 2020). Some estimates suggest that implementing universal health care in the US might save 68,000 lives annually (Galvani, Parpia, Foster, Singer, & Fitzpatrick 2020), and that there are perhaps as many as 155,000 deaths attributable to pollution per year (Sifferlin 2017). According to one estimate, the Iraq War resulted in about 460,000 excess deaths from 2003 to 2011—about 57,500 annually during that period (Hagopian et al. 2013).

I have two notes on premise (2). First, I take it that we have some fairly clear intuitive sense of what it means to “prioritize” issue X over issue Y. In the context of voting, it means something like being prepared to vote for a candidate who will best address X over one who will best address Y (for whatever you think “best addressing” comes down to).¹ In the context of activism, it means something like being prepared to devote effort and resources to issue X over issue Y, at least when you cannot devote all the necessary resources to both. For a politician, it means something like being prepared to put effort, “political capital,” etc. into X rather than Y, at least when you cannot do everything you’d like regarding both. In the context of state action, it means something like addressing issue X through legislation, enforcement, etc., over addressing Y, at least if you cannot do everything you’d like to address both. Etc. (These definitions are too simple to be quite right, but the details don’t matter too much, and again the intuitive idea is fairly clear.) We could also formulate more specific versions of the argument which focus on these specific consequences. For example:

And so on.

A second point is that I want “unless there is some sufficiently good reason not to” to be interpreted in the broadest way possible. Perhaps some other issue is more pressing due to consequentialist considerations besides death, such as causing suffering. Or perhaps it’s more pressing for some non-consequentialist reason: endorsing this premise does not commit one to any particular ethical theory, only to the claim that there is a presumption in favor of prioritizing issues insofar as they result in the unjust killing of more people. This is important, since obviously many pro-life people reject consequentialism, especially those with certain religious backgrounds. “Sufficiently good reasons not to” might also just be practical considerations. Perhaps some issue is really important, but there’s not much we can do about it, or perhaps doing something would be too costly or have bad side effects. Or alternatively, perhaps some other issue isn’t as important, but it would be easy to completely solve it. Or suppose I’m deciding where to focus my activism. Maybe one issue is very important but already gets plenty of attention, so it’s better in terms of marginal impact to focus my efforts on some other, more neglected issue. Understood in this way, premise (2) seems very plausible. It’s largely just an application of the very attractive idea that, in a situation where you can’t save everyone, there is a presumption in favor of saving more people.

Premise (1) is uncontroversial, given the pro-life view, and premise (2), as mentioned, seems very plausible. Premise (3) is self-explanatory given my explanation of premise (2), and it initially also sounds very plausible, given the pro-life view. It initially seems hard to imagine what could be more important than combating the murder of fifty-six million people every year. Of course, there are ways to challenge it. Maybe we can’t really do much in practice to reduce abortion. Maybe we should instead prioritize factory farming because it produces so much suffering, or climate change because it threatens our civilization, etc. Notably, here I’m going to argue that premise (3) is undermined by the Usual Response: it’s what generates the conflict between the Usual Response and Body Count Reasoning. This may have more “bite” against Body Count Reasoners than other challenges, since many pro-life philosophers already accept the Usual Response.

It’s also important to note that there are arguments for prioritizing abortion which don’t superficially have the form of Body Count Reasoning, but which in some way presuppose it or something like it. Calum Miller informs me that he accepts something like the following argument for voting for Donald Trump (the statement of the argument is mine): abortion is relevantly like the Holocaust; if Donald Trump wanted to combat the Holocaust and Joe Biden wanted to allow it, it would be true that you should vote for Trump; but Trump does want to combat abortion while Biden wants to allow it; so you should vote for Trump. This is part of a larger family of pro-life attempts to argue that we should prioritize abortion because it’s relevantly analogous to some historical atrocity which would obviously take priority. (I’ll discuss this in Section 2.2.) We could schematize these in something like the following way (the Holocaust could of course be replaced with some other historical atrocity):

The Argument from Analogy:

Obviously the hard work here is being done by (1*). Even if abortion is wrong, why think it’s “relevantly like” the Holocaust (in the sense of “relevantly like” needed for (3*) to be plausible?) Of course, there are various answers one might give. But we’ll see below (in Section 2.2) that one common answer involves appealing to the claim that far more people are unjustly killed by abortion than were unjustly killed in the Holocaust.² While the argument from analogy does not superficially invoke Body Count Reasoning, these instances of it tacitly presuppose something like it. Specifically, we might say that one way of supporting (1*) is via appeal to something like the following argument:

BCR-Analogy:

We could ask for clarification about some of the premises, but the details really aren’t important. My argument later will also imply that the Usual Response undermines (3a), given a reasonable interpretation of it. And of course there are various other arguments which, while superficially distinct from Body Count Reasoning, may presuppose something like it.

2.2. Examples

Something like Body Count Reasoning plays an important role in pro-life discourse, appearing in media sources, statements from intellectuals, religious leaders, and politicians, and even in laws. (If you already agree with this, feel free to skip over the following examples.) Consider an example from Catholic Answers Magazine, published by Catholic Answers, a popular Catholic apologetics organization and a lay-run apostolate of the Diocese of San Diego. The writer Jimmy Aikin (2004), considering whether other issues might justify “materially cooperating” (see Section 4.2) with the evil of abortion by voting for a pro-choice candidate, clearly invokes Body Count Reasoning:

What kind of reason would be needed to vote for a pro-abort candidate for president? Something unimaginably huge. … Consider: A million and a half new Americans are murdered every year by abortion. … No other issue involves numbers that high. Nothing short of a full-scale nuclear or biological war between well-armed nation states would kill that many people, and we aren’t in imminent danger of having one of those. Not even terrorists with weapons of mass destruction could kill that many people. … Jobs? The economy? Taxes? Education? The environment? Immigration? Forget it. … All of them together cannot provide a reason proportionate to the need to end abortion. Make no mistake: Abortion is the preeminent moral issue of our time. It is the black hole that out-masses every other issue. Presenting any other issues as if they were proportionate to it is nothing but smoke and mirrors.

In response to the same question, Joseph Naumann and Robert Finn (2008), then Archbishop and Bishop (respectively) of Kansas City, gave a similar answer:

Could a voter’s preference for the candidate’s positions on the pursuit of peace, economic policies benefiting the poor, support for universal health care, a more just immigration policy, etc. overcome a candidate’s support for legalized abortion? In such a case, the Catholic voter must ask and answer the question: What could possibly be a proportionate reason for the more than 45 million children killed by abortion in the past 35 years? Personally, we cannot conceive of such a proportionate reason.

The reasoning here seems straightforward: abortion should be prioritized because it kills so many people, and no other other consideration is strong enough to defeat the presumption in favor of prioritizing an issue that kills so many people.

Or consider an article written by the Catholic theologian Thomas D. Williams (2018a) for the right-wing website Breitbart. In the article, which proclaimed abortion the “leading cause of death in 2018,” Williams says that “the staggering number of deaths from abortion, in fact, has led certain observers to call abortion ‘the social justice cause of our time,’ since judging from the sheer magnitude of the problem other human rights issues pale in comparison.” The statement that abortion is the “social justice of our time” due to its death toll is another fairly clear invocation of Body Count Reasoning. In a different article for Breitbart, Williams (2018b), reporting on a study by a group of pro-life researchers (Studnicki, Fisher, & MacKinnon 2016), similarly declares that abortion is the “number one cause of death in the US” and “accounts for a disturbing 61 percent of deaths of African-Americans.”

In response to the same study, an article in the pro-life website LifeNews calls abortion “the leading cause of death, surpassing heart disease and cancer” (Stark 2018). The article quotes the study’s lead researcher, James Studnicki, employing something like Body Count Reasoning as follows:

As a cause of death, the major one for Hispanics and African Americans, abortion would be at the top of the scientific agenda in the U.S., and with a funding priority consistent with its importance. Imagine the urgent scrambling among federal health bureaucracies if some mysterious new virus or bacteria were killing more than a million children each year. (Stark 2018)

Specifically, this seems like a version of the Body Count Reasoning-based Argument from Analogy: a new disease with an equivalent death toll would get prioritized over everything else, but abortion is relevantly like such a disease, therefore, etc. The article also explicitly appeals to the claim that abortion is murder, so that preventing it is more urgent than preventing natural death:

But it’s not just the sheer scale of abortion that separates it from other causes of death. Abortion, unlike natural or accidental death, is intentional killing. Cancer and heart disease are tragic and should be fought with compassion, but they are not injustices. Abortion is an injustice. It is a violation of the right of all human beings to life (the right not to be intentionally killed). (Stark 2018)

At other times, authors attempt to underscore the importance of combating abortion by simply comparing the number of abortions to the number of deaths by other causes, without spelling out the other steps of Body Count Reasoning. Though Body Count Reasoning is less explicit here, something like it must be implicit, or else the point of the comparisons would be unclear. Ronald Reagan (2004) made such a comparison, writing that “the consequences of [Roe v. Wade] are now obvious: since 1973, more than 15 million unborn children have had their lives snuffed out by legalized abortions. That is over ten times the number of Americans lost in all our nation’s wars.”

Another version of something like the Argument from Analogy appears in HB 314, an Alabama law whose implementation is currently blocked as unconstitutional by the Federal courts (Chandler 2019). The bill makes “abortion and attempted abortion felony offenses except in cases where abortion is necessary in order to prevent a serious health risk to the unborn child’s mother” (1). After asserting the equality of “unborn children” and other humans (2–3), the law justifies its unconstituional restriction on abortion by stating:

It is estimated that 6,000,000 Jewish people were murdered in German concentration camps during World War II; 3,000,000 people were executed by Joseph Stalin’s regime in Soviet gulags; 2,500,000 people were murdered during the Chinese “Great Leap Forward” in 1958; 1,500,000 to 3,000,000 people were murdered by the Khmer Rouge in Cambodia during the 1970s; and approximately 1,000,000 people were murdered during the Rwandan genocide in 1994. All of these are widely acknowledged to have been crimes against humanity. By comparison, more than 50 million babies have been aborted in the United States since the Roe decision in 1973, more than three times the number who were killed in German death camps, Chinese purges, Stalin’s gulags, Cambodian killing fields, and the Rwandan genocide combined. (4–5)

As mentioned in Section 2.1, the Holocaust is a particularly common target for comparison. Nicola Beisel and Sarah Lipton-Lubet write that:

The claim that abortion is a holocaust appears in the very first public pronouncements of pro-life activists. … The argument that abortion is a holocaust appealed partly because of the magnitude of death. Pro-life literature routinely cites the number of fetuses who have died since abortion was legalized. (2003: 6)

In 2014, the South Carolina state senator Mark Fair caused controversy by comparing abortion to the Holocaust. When a Democratic senator objected that “when 6 million people, men, women and children, are dragged into gas chambers, I don’t think that’s an appropriate comparison,” Fair responded with: “Six million, it shouldn’t have been one, for crying out loud. Six million is incredibly horrible, what they endured. But those 60 million [the approximate number of abortions since Roe v. Wade] taken out of the womb will never have a chance to be dragged into anything. They’re gone” (Smith 2014).

2.3. Miscarriage

So: Body Count Reasoning is popular. Before moving on, I’ll briefly note a possible worry with Body Count Reasoning which isn’t the one I’ll raise. Contrary to some of the claims cited above, if personhood begins at conception, the leading killer of human persons is probably not induced abortion, but rather natural embryo loss, usually very early in pregnancy. Some authors (e.g., Ord 2008; Berg 2017; Lovering 2017; Simulket 2017; Greasley 2018: 34–37) have claimed that this exposes a problem in the pro-life position. They think it’s absurd to say that the natural loss of recently implanted embryos is a grievous catastrophe, and/or think that the fact that pro-life people don’t seem concerned about miscarriage exposes hypocrisy or intellectual dishonesty on their part. We might frame the argument like this:

The Miscarriage Argument:

Relevant here is that this might also seem to undermine pro-life Body Count Reasoning: maybe body counts really support prioritizing miscarriage over everything else, not induced abortion.

Common responses in favor of prioritizing intentional abortion over miscarriage involve denying that much can feasibly be done to significantly reduce the number of miscarriages and stressing the claim that induced abortion is more morally urgent since it involves, not just death, but killing (e.g., Tollefsen 2008; Dodsworth et al. 2008; Friberg-Fernros 2015; Blackshaw & Rodger 2019; cf. sec. 4.2).³ Both reject premise (3m): one claims that a good reason for prioritizing abortion over miscarriage is that more can be done to reduce abortion, while the other claims that a good reason is that abortion involves unjust killing rather than natural death.

Whether these responses successfully justify pro-life people prioritizing abortion over miscarriage is controversial. Here, I’ll just grant that these responses succeed, since the issue I raise later has nothing to do with miscarriage. I’ll also make two additional points. First, even if consistent Body Count Reasoning does lead to prioritizing miscarriage over everything else, it might also lead to prioritizing abortion over everything else except miscarriage. This would also generate the conflict I discuss later. Second, one variant of Body Count Reasoning employs it to argue, not that reducing abortion is more important than other issues, but rather that it’s important enough to justify the costs imposed by anti-abortion laws. Francis Beckwith employs it like this when discussing deaths among women seeking illegal abortions:

one should not minimize that such deaths were significant losses to the families and loved ones of those who had died. On the other hand, one cannot ignore the more sobering fact that if the unborn is a full-fledged member of the human community, these abortion-related maternal deaths, though tragic, pale in comparison to the nearly 40 million preborn human beings who have been killed by abortion in the United States since 1973. (2007: 122)

Since the comparison here is not between abortion and other issues but between abortion and the costs of outlawing abortion, it’s unaffected by whether miscarriage or abortion is a greater moral problem. But insofar as this reasoning involves appealing to the large number of abortions, it’s affected by the conflict discussed later.

3. The Embryo Rescue Case

The Embryo Rescue Case (ERC) features prominently in the literature on abortion.⁴ While there are variants, the idea is something like this:

Embryo Rescue Case: There is a fire in the IVF clinic. In one room is a child (or baby, or adult); in the other room is a tray with a large number of embryos. You can save the child or the embryos, but not both.

Most people think you should save the already-born person. (It seems safe to say that this won’t change depending on whether we make that person a baby, a child, or an adult, and in fact, different presentations of the case use each of these possibilities.)⁵ Kate Greasley (2018: 27) writes that “The assumption by discussants who invoke ERC is that almost everyone confronted with such a scenario will respond that you must rescue the fully formed baby.” Similarly, Christopher Kaczor (2018: 148) says that “Virtually everyone, even those who call themselves pro-life, think the firefighter should save the girl rather than the embryos.” This might seem to pose a problem for the view that an embryo is a person with the same basic moral status as a developed baby. Greasley (2018: 28) writes that:

When given a choice as to whom to save, and when all other things are equal. … our moral duty is arguably to save the many over the few. … Consequently, if the ERC intuition is correct, this suggests that the moral status of the baby seriously outweighs that of an embryo, such that, even though the embryos are greater in number, saving the baby is the only reasonable course of conduct.

If this line of reasoning works, it might show, not only that the embryo possesses less moral status than a child, but also that it possesses much, much less, since it seems right to save the child over very many embryos. Call the argument for the claim that ERC is a counterexample to the pro-life view the Embryo Rescue Argument.⁶

The Embryo Rescue Argument might be framed like this:

The Embryo Rescue Argument:

As Greasley (2018: 28) notes, since most people find premise (5e) compelling, the “favored response” to the Embryo Rescue Argument grants that you should (or at least may) save the child, but attempts “to explain the ERC intuition in ways that are consistent with the moral equivalence of embryos and human babies.” The two factors I mentioned as potentially undermining premise (3m) in the Miscarriage Argument—intractability and the unjust killing/natural death distinction—don’t apply in ERC. Still, it clearly doesn’t immediately follow from the fact that one should save the child that the child has a greater basic moral status. Perhaps I should save a scientist who is on the verge of curing a terrible disease rather than a larger group of ordinary people, but this doesn’t mean the ordinary people don’t enjoy full personhood. We can even alter ERC so that it seems permissible, or even obligatory, to save the embryos instead (perhaps the embryos will grow up to do great things, while the born person is Hitler). Clearly this wouldn’t show that embryos have greater moral status than born people. While it may be right that “when all other things are equal … our moral duty … is to save the many over the few,” it may be that all other things are usually very unequal between the child and the embryo, despite their having fundamentally equal moral status. This is what I’m calling:

The Usual Response: Premise (3e) is false. Even if embryos are persons, other reasons are sufficient to justify saving the born person. These reasons are more or less those discussed in the next two paragraphs of this very paper.

There are a few candidates for the “reasons” mentioned in the Usual Response.⁷ Dodsworth et al. (2008: 30) note that “In several ways, embryos have much less to lose by death than adult humans do: they do not have life plans, desires, responsibilities, or intentions that are thwarted; early embryos do not feel pain, and they are not yet participating members of any personal relationships.” Robert George and Christopher Tollefsen (2017) write that:

there are differences between the embryos and the five-year-old girl that are or can be morally relevant to the decision concerning whom to rescue. For example, the five-year-old will suffer great terror and pain in the fire, but the embryos will not. Moreover, the family of the five-year-old presumably loves her and has developed bonds of attachment and affection with her that will mean much greater grief in the event of her death than in the event of the death of the embryos. While these concerns would not justify killing, they can play a legitimate role in determining how we may allocate scarce resources and, in some cases, whom we may or should rescue.

And Christopher Kaczor (2018: 149) writes that:

the firefighter in saving the girl rather than the ten human embryos might reasonably conclude, “This girl has plans, goals, dreams, and desires that would be thwarted. She has social relationships with her mom, dad, brothers, sisters, and friends. Her teachers have already taught her to read and invested so much in her education.” By contrast, these circumstantial considerations do not bolster the case for rescuing the human embryos. … In cases of triage, a wide variety of considerations is morally relevant in making a determination about which lives are to be saved, and we need not deny the basic value of any human being in order to come to such decisions.

From these, we can see that there are three or so main factors which allegedly justify saving the child, even if the embryos are persons. First, there is the “pain and terror” the child will feel. A second factor involves the capacity for future-directed attitudes. The child, who has a sense of herself persisting through time, “has plans, goals, dreams, and desires [about the future] that would be thwarted,” whereas the embryo has no such attitudes. A related point, which draws on Jeff McMahan’s (2002: 165–86) influential account of the badness of death, appeals to so-called “time-relative interests” (see, e.g., Liao 2006; Friberg-Fernros 2015). As S. Matthew Liao (2006: 143–44) explains:

it could be argued that the embryos have little or no time-relative interests while the grown child may have very strong time-relative interests. … To have time-relative interests is to be able to stand in some psychological relations to one’s future and past selves. The strength of one’s present time-relative interests depends on how strongly one is psychologically connected to those future and past selves. For example, an infant will typically have weaker time-relative interests in, for example, continuing to live than a grown adult, since an infant has little or no awareness of his or her future self.

The basic idea is that the lack of psychological connectedness with its future self gives the embryo less of an interest in the flourishing of that future self, so that it has less to lose through death. A third factor involves the child’s having had time to develop relationships with other people. If the child dies, these relationships will be tragically cut short, others will experience grief, the hopes and plans of these others for the child will be thwarted, and the investments they’ve made in the child will come to nothing. Perhaps there are other factors which make the death of the child worse, too, but I doubt they will make any important difference to my argument.⁸

Defenders of the Embryo Rescue Argument respond by attempting to modify the case to control for these elements (Greasley 2018: 31). For present purposes, I’m not concerned with who gets the better of this. Obviously, if ERC really is a successful counterexample to the pro-life view, then Body Count Reasoning, insofar as it presupposes the pro-life view, fails. My aim is to point out that the Usual Response undermines Body Count Reasoning, whether or not it succeeds as a response to the Embryo Rescue Argument.

Of course, the Usual Response is not the only possible response to the Embryo Rescue Argument. One could instead reject the intuition that you should save the already-born person. Given how widespread and compelling the intuition is, merely biting the bullet isn’t attractive. The philosopher who employs this approach might instead offer an error theory explaining why the intuition isn’t trustworthy (e.g., Moreland & Rae 2000: 275; Beckwith 2007: 169–70; Hendricks 2019). I won’t attempt to seriously assess this, which would require a deep digression into moral epistemology and other issues. I don’t myself find this response attractive, and obviously most other people don’t, either, given how frequently even pro-life philosophers grant the ordinary judgment about ERC. Further, I will just assume that we should trust our intuitions in cases like ERC when I present some variants of it in Sections 4.2 and 4.3. However, I will note that the conflict between the Usual Response and Body Count Reasoning may pose a problem for the Body Count Reasoner even if the error-theoretic response succeeds. Proponents of the Usual Response think it is independently plausible that the factors they point to can justify saving the already-born person. (Otherwise the Usual Response would be ad hoc.) If this is right, we don’t actually need to directly trust the intuition about who to save in ERC. We might instead appeal to the independent plausibility of the claim that future-directed attitudes, relations with others, etc. can justify saving one person rather than a much larger number of people. This would be enough to generate the problem found in Section 4.1. Whether we could adequately rebut the attempts to resolve this conflict which I discuss in sections 4.2 and 4.3 without (as I do) appealing to our intuitions about cases broadly like ERC is not a question I’ll address.

4.1. Weighing Lives

We can now see the apparent conflict between Body Count Reasoning and the Usual Response: if the latter successfully undermines (3e) in the Embryo Rescue Argument, it may also undermine (3), (3v), and (3a) in the various forms of Body Count Reasoning. If my reasons to save very many fetuses from death tend to be weaker than my reasons to save even one born person from death, it looks much harder to defend the Body Count Reasoner’s claim that the presumption in favor of prioritizing issues which kill more people is undefeated in the case of abortion. I’ll discuss some potential ways of lessening the tension in sections 4.2 and 4.3. But for now I will grant that the Usual Response shows that our reasons to prevent an abortion are much weaker than our reasons to save a born person from death and will ask what follows.

Because there are so many abortions, it could still be that, in the aggregate, combating abortion is extremely urgent. But exactly how urgent it is relative to other issues will depend on how much weaker are our reasons to save fetuses. Suppose, for simplicity, that my reasons to prevent an abortion are always equal in strength to my reasons to save an embryo in ERC. (This is questionable and will be revisited in Section 4.3.) What is the approximate number of embryos such that we should be indifferent between saving them and saving a child? Some people may be inclined to think that, even if the pro-life view is correct, we should not save any number of embryos over an already-born person, perhaps due to anti-aggregationist arguments like those defended by Scanlon (1998: 229–41). In that case, it might not make sense to take any effort away from other life-and-death issues in the name of reducing abortion. Suppose one instead thinks saving a thousand embryos instead of an already-born person is a fair trade. A 1000-to-1 ratio would make preventing the 868,000 abortions which occur annually in the US as urgent as preventing 868 other deaths--slightly less than the number of people shot by police in the US annually (“Fatal Force” 2020). A 100-to-1 ratio makes it equivalent to preventing 8,680 deaths--about an eighth of the number of drug overdoses. A 10-to-1 ratio makes it equivalent to 86,800 deaths--about 25% greater than the number of lethal drug overdoses, and about half the total number of deaths of despair or the estimated number of pollution deaths. Ratios like these would make abortion an important issue, but not overwhelmingly more important than any one of a number of other socio-political issues. Unless abortion should be prioritized for other some other reason unrelated to Body Count Reasoning, the considerations at play in the Usual Response suggest that the case for prioritizing abortion over everything else fails, even granting the pro-life position.

Or perhaps this is too quick. I now consider three potential lines of objection.

4.2. Objections from the Death/Unjust Killing Distinction

One possible way of reconciling Body Count Reasoning with the Usual Response involves emphasizing the supposed status of abortion as unjust killing, not just death. Responses to the Embryo Rescue Argument sometimes note that the fact that we should let the embryos die doesn’t entail that we may permissibly kill the embryos, or, indeed, that there is even any difference in the strength of our reasons not to intentionally kill embryos as compared to children (see fn. 7). This, itself, appears irrelevant to Body Count Reasoning. The question at issue in pro-life voting and activism appears to be, not whether you may, yourself, get an abortion, but rather the strength of your reasons to prevent other people from getting abortions. So at least initially, taking political action against abortion really does seem more analogous to deciding whom to rescue than to deciding whether or whom to kill, so that the analogy with ERC holds.

I will consider shortly whether this prima facie appearance holds up. But first I will consider a different move. Someone could claim that our reasons to rescue embryos from being unjustly killed are much stronger than our reasons merely to rescue them from death—and in fact are much closer to our reasons to save already-born people from death. Perhaps this prevents us from reasoning from the relative weakness of our reasons to save the embryos in ERC to the weakness of our reasons to prevent abortion. (After all, the cause of the fire in ERC is usually left unspecified, but can be presumed accidental, rather than a deliberate act of killing.)

However, this response winds up being about as counterintuitive as saying we should save the embryos in the ordinary case. Consider:

Revised Embryo Rescue Case 1 (RERC1): There are two fires in the IVF clinic. One fire is threatening a child and was started by a lightning strike. The other fire is threatening a tray with a large number of embryos and was started by the woman who provided the eggs which were fertilized to make these embryos. She started the fire with the intention of destroying the embryos. You can save the child or the embryos, but not both.

On the pro-life view, if I fail to save the embryos, they will be unjustly killed by their mother. But I doubt anyone’s reaction to this case will be importantly different than to the ordinary ERC: if you think you should save the child over a certain number of embryos in ERC, you’ll more or less think the same here, even if you agree that the embryos are being unjustly killed in RERC1. So the unjust killing component does not seem to significantly strengthen our reasons to save the embryos.

Someone could claim that RERC1 is a bad analogy because it still involves attempted murder. Perhaps what makes an unjust killing worse than a natural death is the fact that it involves an attempted unjust killing, and the attempt is bad even if, due to luck, it fails. On this view, we might have stronger reasons to preempt a successful murder attempt than to prevent a natural death, but not stronger reasons to thwart a murder attempt already in progress than to prevent a natural death, since in the latter case, the special badness of the attempt obtains either way. And pro-life laws probably might well deter many women from even attempting to procure abortions. But the ethical claim here isn’t intuitive either. Consider:

RERC2: The situation is as in RERC1, except that instead of saving the embryos from a fire in progress, the alternative to saving the child is installing a security measure which will deter the woman from setting the second fire.

Here, if I act to save the embryos, there will be no attempt. But my reasons to save the embryos don’t seem significantly stronger than in RERC1. So the existence of an attempt doesn’t seem to make a big difference. Someone could instead claim that what matters is really not the attempt, but instead something like the willingness of the mother to kill the embryos if she can get away with it. That obtains whatever I do in RERC2. There are two problems here. First, it is unclear whether pro-life legislation could significantly decrease the willingness of people to get abortions, so it’s not clear that this would have any bearing on the strength of our reasons to pursue such legislation. Second, consider:

RERC3: The situation is as in RERC2, except, instead of installing a security measure, I can present the mother with pro-life literature which will prevent her from being willing to kill the embryos.

Here, I can prevent the willingness, but this fact again doesn’t seem to make my reasons to save the embryos appreciably stronger.

A different approach is inspired by work from Thomas Pogge (2010) and by Bruce Blackshaw and Daniel Rodger (2019). Pogge (2010: 126) denies that it is more important to prevent others from being unjustly killed than to prevent them from naturally dying. However, Pogge also argues that, if the pro-life view is correct, all members of society should be regarded as participants in the practice of abortion, since “as citizens of a democratic society, we co-determine, and thus share a special responsibility for, its laws and social institutions,” making us “responsible for helping to bring these deaths about by participating in, maintaining and enforcing a legal system that, by permitting abortions, foreseeably results in these extra deaths” (2010: 127). If this is right (and abortion is unjust killing), our reasons to change abortion laws might be less like our reasons to prevent an unjust killing and more like our reasons not to commit an unjust killing. The latter are widely agreed to be stronger.

Blackshaw and Rodger (2019: 113–15) suggest that this provides some reason for pro-life people to prioritize induced abortion over miscarriage. Similarly, the Body Count Reasoner might claim that we should prioritize abortion at least over other issues in which we are not similarly complicit, even if the death of a fetus tends to be much less bad than the death of a child or adult. But there are two problems with this strategy. First, it would not justify prioritizing abortion over other issues in which (by the same reasoning) we are also all complicit. And an account which is expansive enough to make us all complicit in abortion will probably also make us all complicit in many other things. (Pogge himself goes on to argue that we are similarly all complicit in global poverty, and should prioritize fighting this even if the pro-life view is true.) Second, this again commits us to counterintuitive judgments. Consider:

RERC4: The situation is as in RERC3. Also, it’s legal for the mother to start the fire.

Here again, I doubt that anyone who agrees with saving the child in ERC will think you should instead save the embryos in RERC4. So the legality of abortion, and our alleged attendant participation in it, doesn’t intuitively seem to make a big difference.

Here’s a final move. Perhaps RERC4 illustrates that the mere fact that abortion is legal does not seriously change the calculus. But someone could claim that actually voting for, or otherwise supporting, a pro-choice politician necessarily represents “cooperation” with evil, “help afforded another … to carry out his purpose of sinning” (McHugh & Callan 1929: 616, quoted in Rubio 2011: 104), making it different from a rescue situation: perhaps our reasons not to cooperate with abortion are much stronger than our reasons to rescue embryos from unjust killing. The concept, and morally problematic nature, of cooperation with evil plays an important role in Catholic moral theology, so this maneuver may be particularly appealing to pro-life Catholics.

In evaluating this response, we must first ask what counts as cooperating with evil: do I have to do something—contribute to the wrong act in some way—or is it enough if I merely allow it to happen through “silence or nonresistance” (Rubio 2011: 105)? If merely allowing something to happen when I could stop it counts as cooperation with evil, and if my reasons to avoid cooperation with the unjust killing of a fetus are comparable in strength to my reasons to save a child, then the pro-life person again appears committed to counterintuitive judgments in my RERC cases, since the intuitively correct response involves allowing what the pro-life person views as the unjust killing of very many embryos in order to save just one already born child.

On the other hand, suppose I do actually have to contribute to the evil in order to cooperate with it. This raises at least three issues. First, it isn’t obvious that politicians, in supporting pro-choice laws, themselves actually contribute to abortions, rather than merely allowing them (at least where the issue is the mere legality of abortion, rather than, say, whether it should be publicly funded). Second, this consideration, even if successful on its own terms, has somewhat limited scope: if it justifies single-issue abortion voting, it might not justify prioritizing other sorts of pro-life socio-political involvement. If I spend time, energy, and money on activism aimed at promoting (say) anti-pollution legislation rather than pro-life legislation, it isn’t clear that I materially cooperate with abortion if material cooperation requires contributing to the activity, and so it isn’t clear that this consideration favors one relevant kind of activism over another. Third, even when it comes to voting (or otherwise supporting a political candidate), it isn’t clear that this response justifies being a single-issue abortion voter as opposed to being a non-voter (and otherwise refraining from political involvement). It is often—maybe always—the case that all viable candidates for high office will commit seriously immoral actions if elected. If voting for such a successful candidate constitutes material cooperation with evil, if material cooperation with evil is much worse than allowing evil or failing to bring about good, and if one can avoid such cooperation by refusing to vote, perhaps this account would really support political non-involvement.

But set all that aside. Another, even more serious problem is that the appeal to material cooperation is not supported by the standard account of the ethics of material cooperation anyway. This account, as developed in Catholic moral theology, invokes at least two central distinctions. The first is between formal and merely material cooperation. If I formally cooperate with evil, I cooperate with evildoing for the purpose of bringing about the evil, and this is regarded as always wrong. In the case of merely material cooperation, I cooperate in bringing about the evil, but don’t do so intentionally (even if the evil is a foreseen side-effect of my action) (Rubio 2011: 104–5). Presumably, a pro-life voter won’t support a pro-choice politician because the politician is pro-choice, so the issue is about merely material cooperation. Among types of material cooperation, Catholic moral theology draws a further distinction depending on how closely connected to the evil my own action is. For instance, the editor of a racist tract is more directly involved in the author’s evildoing than is the printing press operator, who is in turn more directly involved than the owner of the electric plant which powered the author’s computer. More remote cooperation is considered easier to morally justify. This is not to say that even very remote cooperation is automatically justified. But whereas very direct material cooperation may be morally indistinguishable from formal cooperation, remote material cooperation can be justified in the presence of a proportionately strong countervailing reason (Rubio 2011: 105).

My action of voting is very remote from any evils to which it contributes—these evils will follow only through the actions of many other agents. So, by the above account, it seems that voting for a pro-choice candidate could be justified, in the presence of a proportional reason for doing so. This was in fact the conclusion reached by Cardinal Joseph Ratzinger (2004), later Pope Benedict, in a letter concerning (among other issues) whether Catholics who support pro-choice politicians should be denied communion:

A Catholic would be guilty of formal cooperation in evil, and so unworthy to present himself for Holy Communion, if he were to deliberately vote for a candidate precisely because of the candidate’s permissive stand on abortion and/or euthanasia. When a Catholic does not share a candidate’s stand in favour of abortion and/or euthanasia, but votes for that candidate for other reasons, it is considered remote material cooperation, which can be permitted in the presence of proportionate reasons.

So the question is whether there are proportionate reasons to justify supporting a pro-choice candidate. Obviously, this may depend on other features of the situation. But here’s the significance for us. The point of Body Count Reasoning was to show that nothing else is proportionate to abortion, that combating it should be prioritized over everything else. I called that into question above, and the notion of material cooperation was invoked by the objector as a way of preserving a version of Body Count Reasoning. But we now see that, on the dominant view, appealing to material cooperation to justify single-issue abortion voting requires that we’ve already established what Body Count Reasoning was trying to show, that is, that nothing else is proportionate to abortion. So appealing to material cooperation in order to Body Count Reasoning begs the question.

4.3. Objections from the Timing of Abortion

Here is a very different objection. The discussion above treats all abortion as relevantly analogous to the destruction of an embryo. However, a pro-life person might claim that preventing the killing of a more developed fetus is more urgent than preventing the killing of a recently fertilized egg. (Even people who are not pro-life often agree that “late-term” abortions are harder to justify.) If there is some point during pregnancy past which our reasons to prevent the fetus from being killed are much stronger than our reasons to prevent an embryo being killed, and if a sufficiently large number of abortions happen past this point, abortion might turn out to be an especially urgent issue after all, even if our reasons to prevent many abortions are comparatively weak. This wouldn’t justify prioritizing abortion as such, but it might justify prioritizing abortion past the relevant point.

The problem with this approach comes in trying to identify a point past which it is both the case that (i) our reasons to prevent abortion (even granting the pro-life view) are comparable in strength to our reasons to prevent a born child dying, and (ii) a very large number of abortions happen. Consider the major factors, mentioned in section three, which are taken to justify saving the child over the embryos in ERC: pain, time-relative and future-directed interests, and relations with others. The traditional scientific view has been that fetuses cannot feel pain until fairly late in pregnancy, perhaps around twenty or twenty-four weeks (Derbyshire & Bockmann 2020: 4). Recently, Stuart Derbyshire and John Bockmann (2020) have argued that fetuses may instead be able to feel pain as early as twelve or thirteen weeks. In the US, more than 91% of abortion are performed at or before thirteen weeks, and only 1.3% are performed at or after twenty-one weeks (Jatlauoi et al. 2018). Given the above-cited figure of 868,000 abortions in the US annually, 1.3% of abortions represents about 11,284, and 9% represents about 78,120. These are extremely significant numbers, but ones on a par with the number of people affected by other important issues, as explained in Section 4.1. (And to the extent that there is a chance of pain being experienced even earlier in pregnancy [cf. Abbate 2015], where significantly larger numbers might be involved, presumably the larger number of individuals potentially affected will be at least to some extent counterbalanced in our considerations by the lower probability of there being any pain at all.) Further, some portion of these are for medical reasons which even many pro-life people will regard as justifying, or at least mitigating, abortion (though, to be clear, the majority are not performed for medical reasons [Kaiser Family Foundation 2019]). Further still, while pain caused by abortion might be very intense and is certainly a matter of serious moral consideration, it also seems reasonable to assume that it is generally not as severe as the more prolonged physical and psychological suffering experienced due to some of the other causes of death associated with these issues (for instance, premature and slow death due to inadequate healthcare). Finally, fetal analgesics are available which may ameliorate pain caused during abortion (Derbyshire & Bockmann 2020: 5–6). Accordingly, insofar as we’re concerned about fetal pain, this might support promoting the use of analgesics rather than combating abortion as such. (The former, after all, would probably generate less political pushback and so be more feasible.)

Now consider the other factors mentioned in Section 3. Even a very late-term embryo will have very little in the way of psychological connectedness to its later self, and so weak time-relative interests. It’s not clear that it can have future-directed attitudes at all, and if it can, these will probably be fewer in number and less important to it than the “plans, goals, dreams, and desires” possessed by a child or adult (though perhaps, sufficiently far in pregnancy, one could appeal to the “ideal, dispositional desire” account defended by Boonin [2002], which might also help explain why you should save the newborn in ERC). A fetus will not have developed “social relationships with her mom, dad, brothers, sisters, and friends,” and, in cases where the mother is seeking an abortion, presumably its loss will not cause others more sorrow than the alternatives.

Another potential reason for being especially opposed to later abortion is if one thinks, with David Boonin (2002: 254–60, 276–78), that post-viability abortions are not supported by bodily autonomy considerations in the same way as earlier abortions. (This is compatible with the pro-life position because one might think these considerations mitigate earlier abortions without justifying them.) This may be plausible, but doesn’t help the Body Count Reasoner. The youngest baby to ever survive was born at twenty-one weeks and four days (Pawlowski 2017), so obviously the number of abortions performed after viability cannot exceed the number, cited above, of those performed at or after twenty-one weeks. (Given that the majority of those performed at or after twenty-one weeks are performed before twenty-three weeks [Epner et al. 1998], when survival rates are still low, it’s probably significantly lower.) Of course, yet another potential reason for especially opposing late-term abortions might be that one thinks early embryos lack full moral status while later fetuses possess it. This may also be plausible, but it’s obviously incompatible with the pro-life view.

So I’m skeptical that this response works. Of the considerations invoked in section three to explain why our reasons to save a child are more urgent than our reasons to save an embryo, some do not seem to apply to fetuses at all, even assuming the pro-life view. Others may apply to some fetuses, but probably not to numbers of fetuses much greater than the numbers of people affected by other issues, and these considerations often seem to apply to a lesser degree even in these cases. If the pro-life view is correct, opposing later-term abortion may well be extremely important, in absolute terms. But again, it will not be particularly more important than many other issues—at least insofar as the considerations at play in Body Count Reasoning are concerned.