Ad description
A poster by the pro-life campaign group Both Lives Matter, which appeared in two locations in Northern Ireland in January 2017, featured the claim "100,000 PEOPLE ARE ALIVE TODAY BECAUSE OF OUR LAWS ON ABORTION. Why change that?". Smaller text stated "Stand with us. BOTHLIVESMATTER.ORG". The ad also featured images of two placards which contained the text "#Both Lives Matter" and "PRO-WOMEN and PRO-LIFE".
Issue
Fourteen complainants challenged whether the claim "100,000 People are alive today because of our laws on abortion" was misleading and could be substantiated.
Response
Both Lives Matter said that they had attempted to estimate the number of people alive in Northern Ireland today because the 1967 Abortion Act introduced in Great Britain had not been introduced in Northern Ireland. They said it was not possible to calculate an exact figure, but 100,000 represented a credible and conservative estimate. They provided a link to a report they had published on their website, which contained their first methodology based on the application of the abortion rate in Scotland to Northern Ireland. They also provided a second methodology based on the application of the abortion ratio in Scotland to Northern Ireland, alongside the application of the England and Wales abortion ratio and the combined Great Britain abortion ratio.
Both Lives Matter believed that anyone viewing the ad would understand that the 100,000 figure was, in statistical terms, an estimate. They believed readers would expect the figure to be a reliable and useful indicator of the true position of the number of people alive because of Northern Ireland’s abortion law. They believed it was highly unlikely that a reasonable reader would have understood the figure to be literal. They drew attention to quotes in the report on their website from Dr Esmond Birnie who said “… there is obviously no absolute certainty about what ‘an alternative Northern Ireland’ would have looked like if the 1967 abortion legislation had been applied here as in GB” and that Both Lives Matter had made “plausible and cautious estimates as to what might have happened”.
Both Lives Matter pointed out that the poster contained a clear URL to their website, where readers could find out Both Lives Matter’s reasoning and substantiation for the 100,000 figure, and where it was open to scrutiny and criticism. They said the poster was uncluttered by small print and referred instead to the URL. They believed that it should be acceptable to refer the reader to a website and not present the justification for the 100,000 figure on the poster itself. They said one of the key aims of the poster was to encourage readers to find out more by reviewing the report on the internet and by using the Twitter hashtag displayed on the poster. They believed the inclusion of the URL and the guidance towards the report was a proportionate balance between their desire to have a clear and stark message and the need to explain their reasoning. However, they said that in future they would be willing amend their advertising to make clear that the figure was an estimate, and based on Scottish abortion figures.
Both Lives Matter said that there were no obvious differences between the population in England and Wales and the population of Northern Ireland, so they had initially assumed an abortion rate as the same as for England and Wales. They said it was common practice for policy in Northern Ireland to be based on data drawn from England and Wales. They acknowledged, however, that it could be argued that Northern Ireland was more conservative and less secular than England and Wales, and therefore turned to Scotland as a more appropriate cultural comparator and to calculate a more conservative estimate. They said Scotland had significant cultural, historical and geographical links to Northern Ireland. They said they both had a large number of self-identified Protestant and Catholic people, and their historical connections dated back to the Plantations, and was reflected today in the Ulster-Scots people and language. However, they pointed out that Census data from 2001 and 2011 showed that religious identification was higher in England and Wales than in Scotland, yet abortion numbers were much higher too. Therefore, they argued that a simple correlation between religious identification of a population and the number of abortions should not be drawn.
In their first methodology, Both Lives Matter used the “abortion rate” – the number of abortions per 1,000 women of reproductive age. They explained that they had initially identified the abortion rate in England and Wales each year from 1967 to 2015 using public records, and applied it to the number of women of childbearing age (15–44) in Northern Ireland each year over the same period. They said that while there were no obvious differences between the population in England and Wales and the population in Northern Ireland, they applied a 25.8% reduction based on the Scottish abortion rate to arrive at a more conservative estimate given that Scotland could be seen as a more appropriate cultural comparator. The 25.8% figure was based on the average abortion rate in Scotland between 2008 and 2012. They then deducted the numbers of abortions in the UK that had been carried out on women who were residents in Northern Ireland based on a published Department of Health (DH) figure for the years 1970 to 2015. They then applied a survival rate up to the age of 50 years of 96.5% to the overall figure. For the year 2016, they estimated that the figure would be the same as 2015.
Both Lives Matter said that their report made clear that the methodology was based on the assumption that law affected culture and vice-versa. The report stated that law played an important role in shaping perceived norms, which were a powerful driver of attitudes and behaviours within society. It stated that in the case of the 1967 Abortion Act in the rest of the UK, the signal sent was that abortion was acceptable far beyond any medical reasons. It explained that as abortion was increasingly seen as more acceptable, more women who may otherwise have rejected abortion as an option would have become more willing to have considered it. The report stated that that could be seen in the initial growth of the abortion rate in England and Wales. They added that their projected figures for the number of abortions carried out in the UK on Northern Ireland residents showed that the Northern Ireland abortion rate remained broadly constant from 1970 to 2015, while both the Scotland abortion rate and the England and Wales abortion rate doubled. They believed that could be explained by the fact that the law which applied to Great Britain was not extended to Northern Ireland.
In their second methodology, Both Lives Matter used the “abortion ratio” – the number of abortions per live births. They applied the Scotland abortion ratio to Northern Ireland and deducted the DH figure for the number of abortions in the UK that had been carried out on women who were residents in Northern Ireland. They explained that the first methodology had assumed that fertility and birth rates were the same in Northern Ireland as elsewhere in the UK, yet they were both in fact significantly higher in Northern Ireland. They calculated that the application of the Scotland abortion ratio to Northern Ireland led to a headline figure of 139,379. They said that even if the figure was substantially reduced by over 25% to allow for cultural differences between Scotland and Northern Ireland it would remain above 100,000. They also believed that the disparity between religious identification in the two nations would be smaller when calculated back through earlier decades.
Both Lives Matter calculated that the application of the England and Wales abortion ratio to Northern Ireland would lead to a headline figure of 247,331, and the application of the overall England, Wales and Scotland abortion ratio to Northern Ireland would lead to a headline figure of 237,806. They said even if the Great Britain figure was reduced by up to 55% to account for the cultural differences between Great Britain and Northern Ireland, the headline figure would remain above 100,000.
Both Lives Matter commented on the further factors they had not taken into account in their methodology in response to points raised by some complainants and the ASA. They said they had not included women in Northern Ireland who would have had an abortion but then had a miscarriage and women in Northern Ireland who had illegal abortions. In those cases, they said there was a lack of credible data to obtain accurate figures and that those factors were unlikely to have a significant impact on the overall estimate. Therefore, even if those factors had been taken into account, it would not invalidate the 100,000 figure.
Both Lives Matter said they had also not taken into account the second generation children – the children of those born to those who would otherwise have been aborted. They had not done so because they wanted to arrive at a conservative and robust estimate. They said that if even a quarter of those people alive as a result of the abortion laws went on to have at least one child, the figure would rise greatly. They further calculated that by using a crude live birth rate of 15 births per 1,000 people then there would be 450 second generation children each year for the last 20 years, meaning that the headline figure should be increased by at least 9,000. They said that the initial deliberate omission of that factor was likely to more than offset any other figures and factors that they had not taken into account.
Assessment
Not upheld
The ASA considered that readers were likely to understand the claim to relate to the number of individuals alive today that were born to mothers who would have aborted their pregnancy if it had been legal to do so. We noted the views of many of the complainants that the claim was presented as an absolute, objective claim and therefore would be understood by readers of the ad to be a precise figure rather than an estimate. However, we considered that 100,000 was a large, round number that readers would typically associate with estimates. Furthermore, readers would appreciate that it was not possible to calculate the precise number of abortions that would have theoretically occurred in Northern Ireland if abortion had been legal over the past 50 years. Accordingly, we considered that readers would expect the claim to be an estimate based on a methodology that showed there was a reasonable probability that around 100,000 individuals were alive today who had been born to mothers who would have aborted their pregnancy if it had been legal to do so.
We also considered that, while technically a person “alive today” could include the children of those born to those who would otherwise have been aborted, in the absence of information to indicate otherwise, readers would expect the figure to be directly based on the number of abortions that would have occurred in Northern Ireland had it been legal. They would not therefore expect the figure to include the second generation children.
We acknowledged the comments from Both Lives Matter that the ad provided a website URL where people could find out more about the methodology used and make up their own minds as to the merits of the methodology. Nevertheless, we considered that many readers who saw the poster would not take the additional step of visiting the website, and Both Lives Matter therefore needed to ensure that those people were not materially misled by the claim.
We assessed the general approach of using both Scotland and England & Wales as comparators. Both Lives Matter explained that while they had initially used Scotland as a comparator in their report because of the perceived similar levels of secularism and conservatism in Scotland and Northern Ireland, they argued that England and Wales could equally be used as a comparator. We acknowledged that Census figures indicated that religious identification was higher in England and Wales (68% in 2011; 77.5% in 2001) than in Scotland (56.4% in 2011; 66.7% in 2001), yet the number of abortions was higher too (by at least 25%). However, the Census figures on religious identification were only available for the past 15 years. Furthermore, we considered there was a common distinction between religious identification and religious practise, and that attitudes to abortion were likely to be informed by the religious beliefs of those who practised compared to those who simply identified with a religion. A 2007 report on church attendance figures found that 18% of people in Scotland were regular churchgoers compared to 14% in England and 12% in Wales. A second set of published church attendance figures showed that for every five years, from 1980 to 2015, Scotland consistently had a higher percentage of the population who attended church compared to England and Wales. For instance, in 1980, 17% of the Scottish population attended church compared to 11% in England and 14% in Wales. In 2015, 9% of the Scottish population attended church compared to 5% of the population in both England and Wales. We considered that it was reasonable to assume that religiosity was in part responsible for the lower abortion rate in Scotland and agreed with the initial decision of Both Lives Matter to select Scotland as the comparator. As such, we considered that the figures from Scotland were the most appropriate to use.
We took expert advice on the two methodologies. We acknowledged that all the figures used in both methodologies were taken from reliable and reputable public sources. We were concerned, however, that instead of applying the Scottish abortion rate per 1,000 women each year in the first methodology, Both Lives Matter had instead based the Scottish abortion rate on the average of the annual abortion rates between 2008 and 2012. No valid reason had been provided for using that approach when the abortion rate in Scotland from 1969 to 2007 and from 2013 to 2015 was available from the same report. Furthermore, with the exception of two other years, the years which Both Lives Matter had selected had higher abortion rates in Scotland than every other year. We calculated that using that approach had inflated the ’100,000’ figure by just over 8,500, leaving a figure of 91,488.
In the second methodology, Both Lives Matter had applied the Scotland abortion ratio to the annual number of births in Northern Ireland. The abortion ratio represented the annual number of abortions per the annual number of births, and so additionally factored in the birth rates of both nations. We considered that, notwithstanding any societal differences between the two sets of populations that could affect the number of abortions, it was acceptable in principle to make an estimate based on the application of the abortion ratio. However, we considered that there needed to be an adjustment to account for the smaller total family size that would likely have prevailed were abortion to have been permitted in Northern Ireland. In other words, Both Lives Matter had not factored in that the number of births would have been less in Northern Ireland had abortion been legal because a significant number of those births would have likely been aborted. After subtracting the number of women in Northern Ireland who travelled to the UK to have an abortion and repeating the two further adjustments that Both Lives Matter made in their first methodology to factor in death rates and the 2016 projection, we calculated that, compared to Both Lives Matter’s figure of 139,379, the methodology produced a figure of 106,144.
We noted Both Lives Matter’s argument that the law played an important role in shaping perceived cultural norms, and therefore if abortion had been legalised in Northern Ireland, it might have normalised cultural attitudes to abortion and led to a rate similar to that of Scotland. However, we considered that the different abortion rates between Scotland and England & Wales showed that there could be marked difference between the abortions rates of two countries when abortion had been legalised at the time.
We considered that, notwithstanding any other societal factors, it was reasonable to assume that the more religious a population was, the lower the number of abortions that would occur. The 2007 church attendance figures mentioned above showed that 45% of people in Northern Ireland were regular churchgoers, compared to 18% for Scotland. That suggested that in 2007 the Northern Ireland population was significantly more religious than Scotland. The Census figures also indicated that religious identification was demonstrably higher in Northern Ireland (83.1% in 2011; 86.2% in 2001) than Scotland (56.4% in 2011; 66.7% in 2001) in recent years. Therefore, in the absence of any other societal differences between the two populations, it was likely that the number of abortions in Northern Ireland would have been lower than Scotland even if the 1967 Abortion Act had been introduced there at the same time as Scotland.
Nevertheless, there were various other social, cultural, economic and geographical differences between Scotland and Northern Ireland that could have led to more or fewer abortions in Northern Ireland had it been legal. We considered that there was no quantifiable means of measuring to what extent religious differences and those other societal differences between the two nations would have had on the Northern Ireland abortion figures. While the difference between abortion numbers in Scotland and England & Wales was sizeable, there was no reason to suggest that the same difference would be replicated between Scotland and Northern Ireland. Using Both Lives Matter’s second methodology, a reduction of 6,000 would still leave the figure above 100,000, while a reduction of a few thousand more would leave the headline figure close enough to 100,000 that it was unlikely to materially mislead readers in the context of an estimate.
We noted that there were several other factors that had not been taken into account in the two methodologies. For instance, Both Lives Matter had not factored in illegal abortions that had taken place in Northern Ireland and women from Northern Ireland who had abortions in England but did not give a Northern Ireland address. On the other hand, Both Lives Matter had also not taken into account women who gave birth to twins or triplets, which would have increased the headline figure. We considered that such factors were unlikely to materially affect the overall headline figure, and in some cases it would be hard to find credible data. Because the claim was likely to be interpreted as an estimate, we were satisfied that it was not necessary to have taken into account those factors. We also identified some minor omissions in the data (e.g. the figure obtained from the DH did not include the year 1969), but we were again satisfied that they were unlikely to materially affect the overall headline figure.
In conclusion, we considered that readers would expect from the ad that there was a reasonable probability that there were around 100,000 individuals alive in Northern Ireland today who were born to mothers who would have aborted their pregnancy if it had been legal to do so. While Both Lives Matter’s first methodology had inflated the overall headline figure by just over 8,500, their application of the Scotland abortion ratio – subject to our revisions – produced a figure of just over 106,000. While the calculations did not take account of societal differences between the Northern Irish and Scottish populations, most notably religious differences, there was no evidence on which to suggest that the actual figure would have been significantly lower than 100,000. On balance, we concluded that the evidence indicated that there was a reasonable probability that around 100,000 people were alive in Northern Ireland today who would have otherwise been aborted had it been legal to do so. Because we considered that readers would understand the figure to represent an estimate, we concluded that the claim was unlikely to materially mislead readers.
We investigated the ad under CAP Code (Edition 12) rules 3.1 (Misleading advertising) 3.7 (Substantiation), and 3.9 (Qualification), but did not find it in breach.
Action
No further action necessary.
