The Guardian recently published an article, which I linked to and excerpted here, by George Monbiot claiming a clear link between sex education and contraception use (two of the things that Catholics say they hate) and a reduction in abortion rates (another thing that Catholics say they hate double extra bad), thereby arguing that the Catholic church in opposing contraception and sex education are directly contributing to increases in abortion rates.

Today, Michaela Aston (a spokesperson for Life) is given the opportunity to respond.

This is her first line of attack:

Monbiot is particularly misinformed when he quotes a study which reported that “Rising contraceptive use results in reduced abortion incidence” – over 60% of women having abortions in Britain report that they had been using contraceptives. The journal Obstetrics and Gynecology recently published a meta-analysis of 23 research articles examining the impact of increased access to emergency birth control on unwanted pregnancy and abortions, concluding: “To date, no study has shown that increased access to this method reduces unintended pregnancy or abortion rates.”

What Monbiot actually said: “But, as a study published in the journal International Family Planning Perspectives shows, once the birth rate stabilises, contraceptive use continues to increase and the abortion rate falls. In this case one trend causes the other: ‘Rising contraceptive use results in reduced abortion incidence.’ The rate of abortion falls once 80% of the population is using effective contraception.

Aston does nothing here to demonstrate that the study Monbiot cited was wrong. She does not examine or critique the study at all. She seems merely to claim that the statistics she cited contradict its conclusions. But they don’t!

First, “over 60% of women having abortions in Britain report that they had been using contraceptives” – but even taking that on trust*, it takes nothing away from the assertion made in Monbiot’s study. The fact that 60% of women who decide to terminate an unwanted pregnancy ended up pregnant because of a contraceptive failure – perhaps due to ignorant or lazy or careless (rather than “effective”) use of contraception, which is a trap many of us have fallen into – doesn’t say anything at all about the validity of Monbiot’s assertion unless you also study the rates of effective contraception use in the population generally and the overall abortion rates – the manner in which women having abortions got pregnant has little to do with it.

(* And I don’t think we should. Women seeking to persuade a doctor that they “deserve” or “need” an abortion are quite likely to say that they were using contraception and that it failed, either out of simple shame / embarrassment, or in a deliberate attempt to seem like a more suitable candidate. I know I did, even though in reality the problem was that I used contraception carelessly, not that it failed. So even if 60% of women seeking an abortion do blame contraceptive failure, a massive grain of salt is needed.)

Second, “a meta-analysis of 23 research articles examining the impact of increased access to emergency birth control on unwanted pregnancy and abortions, [concluded]: ‘To date, no study has shown that increased access to this method reduces unintended pregnancy or abortion rates.'” And your point is? For one thing, without knowing whether any study was adequately designed to demonstrate the effect of increased access to this type of contraception, the fact that no study has shown any particular effect is fairly meaningless. In any case, Monbiot’s study was not making any claims about the effect of unspecified increases in access to a particular type of contraception* on abortion rates – the claim was that abortion rates decline when 80% of the population is using contraception effectively. That is a totally different claim and the evidence cited by Aston can have no bearing upon it.

(* And after-the-event contraception is probably the worst type to select as representative of contraception generally, since “effective contraception use” ordinarily involves effective use at the time of having sex, with morning-after pills being strictly for situations where this didn’t happen or where it becomes clear that the contraception used at the time failed, e.g. burst condom.)

The next criticism is this:

Monbiot claims there is “a clear relationship between sex education and falling rates of unintended pregnancy”. There is not. Most papers (including the one in the British Medical Journal that Monbiot cites) find that sex education programmes have little or no impact on rates of teenage pregnancy or abortion. Sweden’s programmes in sex education, and promotion of contraceptives, have been an admired model – yet total abortion rates* there are now higher than ours.

(*Unfortunately Aston does not explain what she means by “total abortion rates” or where she gets her figures. More on this below.)

What Monbiot actually said: “There is also a clear relationship between sex education and falling rates of unintended pregnancy. A report by the United Nations agency Unicef notes that in the Netherlands, which has the world’s lowest abortion rate, a sharp reduction in unwanted teenage pregnancies was caused by ‘the combination of a relatively inclusive society with more open attitudes towards sex and sex education, including contraception’. By contrast, in the US and UK, which have the developed world’s highest teenage pregnancy rates, ‘contraceptive advice and services may be formally available, but in a ‘closed’ atmosphere of embarrassment and secrecy’.”

Sweden does appear to have a good programme of sex education which includes contraception. This appears to have a very beneficial effect comapred with the UK in terms of teenage pregnancies, although abortion rates are still more or less equivalent to ours. Some facts:

  • Sweden’s abortion rate for the last many years has been in the region of 30,000 to 35,000 or so in a population which has now just reached 9 million (as a total rate that is about 20-25% of total pregnancies, varying from year to year). (Source.)
  • At the same time abortions in England & Wales run at about 185,000 to 195,000 or so in a population of about 50-odd million – for 2006 that translated to 18.3 abortions per 1000 women aged 15 to 44 . (Source.)
  • According to my sums, the rate of abortion in Sweden is more or less the same as the rate in the UK. I couldn’t find any direct comparison, and Aston did not say what method of comparison she was using or provide a source. However, when calculating the crude abortion rate (abortions per capita), the results are consistently pretty similar, coming out either a little higher or a little lower depending on the approach I took.

It seems to follow from the above that, despite their better programme of sex education, the Swedish abortion rate is indeed still about the same. Why?

Well, let’s look at those teenage pregnancies that Aston chose to ignore. (Source.)

  • In the mid-1990s (and I suspect things haven’t changed significantly since then) Sweden had a teenage (aged 15 to 19) pregnancy rate of 25 per thousand leading to 17.2 abortions and 7.8 births per thousand.
  • Meanwhile, Great Britain had a teenage pregnancy rate of 46.7, leading to 18.4 abortions and 28.3 births per thousand.

Evidently, although the total number of teenage abortions in Sweden and Britain may have been similar, the total number of pregnancies and births for teenaged women was *much* higher in Britain. The programme of sex education is therefore leading to far fewer pregnancies (and therefore presumably far more effective contraceptive use) and the issue is merely that in Sweden 69% of teenage pregnancies led to abortion, while in the UK only 39% of teenage pregnancies ended in abortion.

My reading of this would therefore be not that widely available sex education makes no difference – my reading would be that widely available sex education leads to better contraceptive choices and fewer teenage pregnancies, so that even when a high proportion of teenage pregnancies end in abortion the overall rate is still comparable with other countries that have rubbish sex education. My reading of this would be that even more and even better sex education is needed to improve contraceptive practices (including abstention!) further, to reduce rates of unwanted pregnancy yet further. There is only so much higher that the proportion of unwanted pregnancies ending in abortion can climb, yet I would suggest that there is still even in Sweden plenty of scope for reduction in the number of teenaged women who become pregnant by mistake.

(In this regard, for US readers, it is worth noting that in the US only 35% of the 83.6 teenage pregnancies per thousand ended in abortion, although because of the higher rate of teenage pregnancy this still resulted in the highest overall rate of abortion at 29.2 abortions per thousand teenaged women.)

So one thing that we can conclude from this analysis is that a simple look at sex education programmes compared with abortion rates is too simple. A comparison of the UK abortion rate with the rate in Sweden – where a very secular and no-nonsense approach to abortion, allowing abortion for free on request, means that it is a more popular and less stigmatised option – is unfair and decidedly does not show that sex education has little or no impact on abortion rates. There are too many confounding factors which, when explored, seem to suggest instead that Sweden are indeed on the right track – their programme has clearly been effective in reducing teenage pregnancies, even if the incidence of abortion has yet to follow suit. The point is that their programme as a whole is also effective in making abortion easily available, with the result that when accidental pregnancy does occur it is a significantly more popular choice there than here.

In any case, Monbiot points not to Sweden as his model but to the Netherlands, where the sex education programme is also extremely comprehensive. In the Netherlands in 2002, the number of abortions per 1000 women aged 15-44 was – compared with the UK rate for 2006 of 18.3 – a mere 8.7. In the same year, the rate of abortions as a percentage of all pregnancies was – compared with 20-25% in Sweden – just 12.7%. And, apparently, these figures include abortions performed on foreign women having an abortion in the Netherlands – we can reasonably speculate that the rate for Dutch women, who have been through the Dutch sex ed programme, is likely to be even lower. (Source.)

Clearly, sex education is working in the Netherlands even if, as we have seen, the picture is more complex in Sweden. It isn’t because the Netherlands has a more restrictive approach to abortion, because they don’t. Could it be that the Netherlands has better rates because it has even better sex education and even lower rates of unwanted pregnancy than Sweden has managed to date? This article states (according to Google’s extract – I don’t have access to the full article) that unwanted pregnancies are rare, although when they do happen they are (as in Sweden) likely to end in abortion; and that the rate of teenage pregnancy in the 1990s – when Sweden and the UK had respective rates of 25 and 46.3 pregnancies per 1000 teenaged women – the rate in the Netherlands was just 9.2. Can Aston answer that? She can’t.

Next, she tries this approach:

In contrast, there is clear evidence that even modest legal restrictions can help to cut abortion rates. Some US states now require parental notification before minors can get abortions. This has led to lower underage abortion rates without necessarily increasing underage births.

Firstly, this doesn’t in fact contrast with or even follow from the previous paragraph, where we weren’t actually talking about about whether legal restrictions affect underage abortion rates. However, Monbiot does talk about this, as it happens, and says: “Chillingly, as the Lancet paper shows, there is no relationship between the legality and the incidence of abortion.”

Without more details of either piece of evidence it is difficult to comment further, but I can say a couple of things. Monbiot cites an actual academic paper showing that there is no relationship between legalisation and abortion rates. He might also have cited the WHO study referred to in this article, which reached the same conclusion. Meanwhile, Aston makes some vague claim that there is, somewhere, possibly, some evidence showing that fewer underage girls get abortions if they have to tell their parents about it, coupled with the vague assertion that this doesn’t necessarily (but, by implication, it may) result in more underage births. Again, reasoning from dubious evidence about one specific aspect of abortion to a general claim about abortion in the round. I know which I find more convincing.

What’s next?

Referring to a Lancet article, Monbiot says: “When the figures are broken down, it becomes clear that, apart from the former Soviet Union, abortion is highest in conservative and religious societies.” The big problem with Monbiot’s argument is that the abortion rates from more religious countries are generally based on conjectural estimates of illegal abortions, and there is a long history of pro-abortion groups deliberately inflating such estimates. The case of Mexico provides a good example. Monbiot cites an estimate of 25 abortions per 1,000 women for Central America. Applying this figure to Mexico suggests about three quarters of a million abortions each year. In fact we now have real data for this country – due to the legalisation of abortion in Mexico City last year – which makes it highly unlikely that there were more than 70,000 illegal abortions a year. This equates to 2.1 per 1,000 women – one tenth of that quoted by Monbiot and far less than in countries where abortion is legal.

In Ireland abortion is illegal and contraception has (at least until recently) been much harder to access than in the UK. Based on the numbers of Irish women having abortions in the UK, their abortion rate is about one third that of England, and there is no evidence of significant numbers of illegal abortions.

Aston is fair to point out that estimates of illegal abortions can only ever be estimates, and that the figures for illegal abortions are less reliable than the figures for legal abortions. It may also be fair to point out that some pro-choicers have in the past been guilty of inflating estimates to make things look worse than they are (at least, I have heard about examples of this) – just as some anti-choicers have been guilty of making unfair estimates or manipulating the available information in support of their own position. Indeed, I suspect that this is something Aston is in fact doing in this article when she discusses the cited rate of illegal abortion in Central America.

This comprehensive resource* published in International Family Planning Perspectives (link courtesy of the Guttmacher Institute) gives some statistics for Mexico, the particular country that Aston has chosen to highlight as a good example (presumably good in the sense that it best supports her own position). The figures given are total abortions of about 500,000 (tolerance 300,000 to 750,000) which works out at a rate of 25.1 abortions per 1000 women aged 15 to 44 or 17.1% of total pregnancies. This is clearly a perfectly plausible estimate – it is in very much the same ballpark as the figures for Sweden and the UK, discussed above.

(* Reliability. These figures are based on a “special study” given the acknowledged difficulties of collecting reliable data on illegal activities. Although there are copious footnotes I couldn’t be bothered to wade through them to get a specific cite for the methodology or authors of the study – and I’m actually quite happy to take the figures on trust as being from a reputable, peer-reviewed source and so likely to be at least half way reliable in the absence of any convincing evidence to the contrary.)

Aston suggests that applying the rate of 25 abortions per thousand women leads to about 750,000 abortions per year – that is the highest end of the estimates given in the resource cited above, but it is worth noting that the rate of 25.1 abortions per thousand was in fact based there on a total number closer to 500,000. So it looks like Aston has got her sums wrong on that one (let’s be charitable). In any case, mentioning a raw figure is fairly meaningless without noting the total population for the country. Mexico has a total population of over 100 million, so those figures are far from astonishing, given than Sweden has 30,000+ abortions for a population of under 10 million and we have around 200,000 abortions for a population of about 60 million.

Aston also suggests that post-legalisation figures for the rate of legal abortion in Mexico, at 70,000 rather than the “750,000” she quotes as the amount allegedly put forward by pro-choicers, show just how completely bonkers the estimates of illegal abortion really were. Firstly, of course, her inflation of the estimates of illegal abortion make the position seem more extreme than it really is. Moreover, her suggestion that 70,000 is a more realistic figure given the experience since legalisation is shockingly misleading. Abortion in Mexico has only been legal since April last year and even now it is only legal for the first three months for people who live in or can travel to Mexico City. There is no suggestion that abortion is to be subsidised by the state, so presumably it will still be out of reach for the majority of women who cannot afford to pay for an abortion. All this means that there are going to be many, many illegal abortions in Mexico even now – those who cannot afford to pay, those who cannot travel to Mexico City, and to a lesser extent those who require an abortion after the three month deadline. As such it is difficult to see how statistics on the rate of legal abortion in Mexico City in the months since legalisation could possibly give us any real indication of the total numbers of abortion throughout Mexico. If Aston’s offered estimates aren’t misleading, I don’t know what is! Monbiot’s figures are much more credible.

Finally, I just wanted to comment briefly on the comparison that Aston makes between Ireland and the UK. She guesstimates that the rate of abortion in Ireland is about a third the UK rate, based on the number of women who travel from Ireland to the UK for an abortion. I haven’t investigated that claim thoroughly, so I’m going to take her at her word (although I would like to ask her more about the rate of illegal abortions in Ireland or how many women travel to other countries for abortion).

But in any case there is one totally obvious reason why it doesn’t help her case even if her claim is more or less fair, that the rate of abortions in Ireland is actually lower. For that to be in some way relevant to the question whether legalisation affects abortion rates, she would also have to claim that “if legalisation made no difference, abortion rates would be the same in all countries irrespective of legal restrictions”. However, that is clearly not the case. Abortion rates are dependent on a huge number of variables including the rates of unwanted pregnancy, cultural attitudes to abortion and the availability of good sex education and affordable contraception. Nobody suggests that abortion rates do not vary between countries or that there is a consistent pattern whereby *all* countries were abortion is illegal have higher rates than *all* countries where abortion is legal. Indeed, Monbiot expressly points out that the UK is a blip with a much higher rate than those of its European neighbours who have legalised abortion. For these reasons, pointing out that Ireland, where abortion is illegal, may have a lower rate of abortions than the UK, where abortion is legal, is completely meaningless.

Moving on, Aston criticises Monbiot’s claim that women are “condemned to death” by the outlawing of abortion. This is where her distortion becomes actually laughable.

She says:

Monbiot claims that, with his stance against abortion, the Pope “condemns women to death”. In the same Lancet issue he referred to earlier, another article gives the latest estimates of maternal mortality rates. Ireland comes out best in the world with a rate of 1 death per 100,000, vastly superior to countries where abortion is legal such as the US (11 per 100,000) and the UK (8 per 100,000).

She suggests that because Ireland has better mortality rates for pregnant women than other developed countries, this puts the lie to Monbiot’s claim that criminalising abortion increases the use of unsafe illegal abortion and hence increases the risk of pregnant women dying, as a result of unsafe abortions. This is absurd, and here’s why.

Firstly, Monbiot specifically makes clear that the women who are “condemned to death” are not, in general, those in developed countries like Ireland. He says: “But the suffering [the Catholic] church causes in the rich nations doesn’t compare to the misery inflicted on the poor… Women with no access to contraceptives will try to terminate unwanted pregnancies. A World Health Organisation report shows that almost half the world’s abortions are unauthorised and unsafe. In East Africa and Latin America, where religious conservatives ensure that terminations remain illegal, they account for almost all abortions. Methods include drinking turpentine or bleach, shoving sticks or coathangers into the uterus, and pummelling the abdomen, which often causes the uterus to burst, killing the patient. The WHO estimates that between 65,000 and 70,000 women die as a result of illegal abortions every year, while 5 million suffer severe complications.” [My emphasis.]

Second, even if Monbiot had not said this then it should have been completely obvious to Aston that Ireland was not going to be a fair comparator. For one thing, she has already claimed that there is a relatively low rate of abortion generally among Irish women, and no evidence of significant illegal abortions in Ireland, since most women who need an abortion are able to travel to UK or elsewhere. On that basis alone it is hardly surprising if there are few or even no maternal deaths in Ireland as a result of illegal abortions. Secondly, even to the extent that there is illegal abortion in Ireland, is it really likely to involve the dangerous methods described by Monbiot? Ireland where women have access to health information, the internet, mail order abortion drugs, and quite possibly clandestine but nevertheless fairly hygienic abortion clinics is hardly the typical setting for dangerous illegal abortions, is it?

All in all, the fact that criminalisation makes no difference to maternal mortality in that particular jurisdiction (even if its small population didn’t make it a statistically less significant country in the first place) is, actually, very unsurprising. And that takes nothing away from the evidence of maternal mortality in other countries. Irish women are protected by the fact that they are part of a developed world that generally has a more liberal approach than their own country; not as Aston would like us to believe, by the very restrictions they face in their own country.

Oh, and in case you missed it the first time, I will repeat Monbiot’s cited statistics on maternal mortality: The WHO estimates that between 65,000 and 70,000 women die as a result of illegal abortions every year, while 5 million suffer severe complications. In the face of those statistics, if the only way you can persuade yourself that illegal abortion is unproblematic is to look at one small country with a developed healthcare system and access to legal abortion in a neighbouring jurisdiction then, Aston, you have a problem.