‘Marches for life’ take place every year in several big cities in Europe (e.g., Paris, Prague, and Warsaw) and Northern America (e.g., Ottawa, San Francisco, and Washington). In France, for example, 11,000 to 45,000 people demonstrated against abortion on January 25, 2015.Parisien, ‘Plusieurs milliers de militants défilent contre l’euthanasie et l’IVG’, Retrieved May 3rd, 2017 at /www.leparisien.fr/laparisienne/societe/paris-plusieurs-milliers-de-militants-defilent-contre-l-euthanasie-et-l-ivg-25-01-2015-4477347.php. Among them, at least five bishops of the Roman Catholic Church were present, as Pope Francis had, in 2014, expressed support for the demonstrators. These demonstrations and the counter-demonstrations that take place simultaneously or at other momentsExpress, ‘Un millier de manifestants pour le droit des femmes à Paris [One thousand people demonstrating for women’s rights in Paris]’, Retrieved May 3rd, 2017 at www.lexpress.fr/actualites/1/actualite/loi-veil-un-millier-de-manifestants-pour-le-droit-des-femmes-a-paris_1641805.html10. seem to capture the essence of public views about abortion, either strongly pro-life or strongly pro-choice.

However, such polarization is not found in empirical research on people’s views of abortion. Most people are sensitive to the circumstances (e.g., the fetal age or the reason for the abortion) in which the abortion decisions are made.J.H. Evans, & K. Hudson, ‘Religion and reproductive genetics: Beyond views of embryonic life?’, Journal for the Scientific Study of Religion 46 (2007): 565-581. M.D.R Evans, J. Kelley & E.D. Zanjani. ‘The ethics of gene therapy and abortion: Public opinion’, Fetal Diagnostic Therapy 20 (2005): 223-234. M.T. Munoz Sastre, C. Pecarisi, E. Legrain, E., Mullet & P.C. Sorum, ‘Acceptability in France of induced abortion for adolescents’, American Journal of Bioethics 7 (2007): 26-32. T. Palermo, M.I. Erazo, V. Hurtado Pinochet, ‘Women’s opinions on the legalisation of abortion in Chile 2009-2013’, Culture, Health & Sexuality 17(2015): 873-890. K.S. Wilson, S.G. García, C. Díaz Olavarrieta, A.Villalobos Hernández, J. Valencia Rodríguez, P. Sanhueza Smith, C. Burks, ‘Public opinion on abortion in Mexico City after the Landmark Reform’, Studies in Family Planning 42(2011): 175-182. Only a minority of participants in these studies expressed an extreme position. For example, in a study conducted in France,M.T. Munoz Sastre, C. Petitfils, P.C. Sorum & E. Mullet, ‘A mapping of the positions of adults in Toulouse, France, regarding induced abortion’, European Journal of Contraception and Reproductive Health Care 20(2015): 158-169. five qualitatively different positions about abortion were found. 23% of participants – who comprised of most of the obstetricians included in the study – responded in a way that was consistent with French current law.Legifrance , ‘Loi n° 75-17 du 17 janvier 1975 relative à l’interruption volontaire de la grossesse [Law 75-17 on induced abortion]’, Retrieved on March 18th, 2017 from www.legifrance.gouv.fr/affichTexte.do?cidTexte=JORFTEXT000000700230&dateTexte=20130913. Among the other participants, 1% considered that the decision to have an abortion is always inacceptable – the pro-life position, 32% considered that the decision to have an abortion is always acceptable – the pro-choice position, and 45% considered that the decision to have an abortion is always acceptable in case of rape, endangered mother’s health or congenital malformation but  never acceptable in case of personal convenience or socio-economic problems. 

Most studies on people’s attitude toward abortion have documented the effect of religious affiliation and/or religious involvement on people’s views regarding induced abortion.M.T. Munoz Sastre, C. Pecarisi, E. Legrain, E., Mullet & P.C. Sorum, ‘Acceptability in France of induced abortion for adolescents’, American Journal of Bioethics 7 (2007): 26-32. J.P. Bartkowski, A.I. Ramos Wada, C.G. Ellison & G.A Acevedo, ‘Faith, race-ethnicity, and public policy preferences: Religious schemas and abortion attitudes among US Latinos’, Journal for the Scientific Study of Religion 51(2012): 343-358. S.G. García, C. Tatum, D. Becker, K.A. Swanson, K. Lockwood & C. Ellertson, ‘Policy implications of a national public opinion survey on abortion in Mexico’, Reproductive Health Matters 12 (2004, Supplement): 65-74. J.A. Hess & J.D. Rueb, J.D., ‘Attitudes toward abortion, religion, and Party affiliation among college students’, Current Psychology 24(2005): 24-42. J. Strickler & N.L. Danigelis, N.L., ‘Changing frameworks in attitudes toward abortion’, Sociological Forum 17(2002): 187-201. E.A. Yam, I. Dries-Daffner & S.G. García, ‘Abortion opinion research in Latin America and the Caribbean: A review of the Literature’, Studies in Family Planning 37(2006), 225-240. For example, Ogland and VeronaC.P. Ogland & A.P. Verona, A.P., ‘Religion and attitude towards abortion and abortion policy in Brazil’, Journal for the Scientific Study of Religion 50(2011): 812-821. reported the results from a survey on people’s attitude towards legalisation of abortion policies conducted in 2002 in Brazil. Non-believers in God were much more in favour of abortion legalisation than Pentecostal or Catholic regular attendees.

The Present Study

The present study examined closely the impact of religious affiliation on people’s personal position on abortion. It used the same material as the one used by Munoz Sastre et al. and was also conducted in France. Three very different religious groups were included in the sample: a group of self-identified Evangelicals, a group of self-identified Catholics, and a group of people who declared that they do not believe in any God.

Although Evangelicals are not very numerous in France, a group of Evangelicals was included in the study because their number is quickly growing,Express, ‘Le nombre d’églises évangéliques en France a triplé en 45 ans [Three times more Evangelical churches in France since 45 years]’, Retrieved May 3rd, 2017 at www.lexpress.fr/actualites/1/styles/le-nombre-d-eglises-evangeliques-en-france-a-triple-en-45-ans_1642809.html. their level of religious practice is higher than that of the other religious affiliations, and they tend to adhere more strictly to the positions advocated by their leaders than most other groups do.B.G. Smith & B. Johnson, B., ‘The liberalization of young Evangelicals: A research note’, Journal for the Scientific Study of Religion 49(2010): 351-360. The National Association of Evangelicals ‘actively, ardently and unwaveringly opposes abortion on demand.’National Association of Evangelicals, ‘Theology of sex’, Retrieved March 18th, 2017 at http://nae.net/theology-of-sex/.

Historically Roman Catholics have been the dominant religious group in France. The Holy See also opposes induced abortion. Canon 1398 provides that conspirators of induced abortion – the woman, her agreeing partner, and the health professionals involved in the process – incur automatic excommunication. Canon 1323 provides, however, that those who are not yet 16, or who were forced to abort, or who acted out of grave fear do not incur excommunication.Code of Canon Law, Retrieved October 8th, 2015 at

www.vatican.va/archive/ENG1104/__P57.HTM. Contrary to Evangelicals, Catholics in France are vanishing in numbers, have a low level of practice and tend not to adhere the positions on sexuality (e.g., regarding contraception) taken by their leaders. Atheists are now the more numerous ‘religious’ group in France: a majority of French adults say they don’t’ believe in any God.CLES, ‘La majorité des Français se passent de Dieu [A majority of French people don’t believe in God]’, Retrieved March 18th, 2017 at www.cles.com/enquetes/article/la-majorite-des-francais-se-passent-de-dieu.

Under French law, abortion is permitted (a) on request until the end of the tenth week of gestation and (b) at later stages for medical reasons attested by two physicians. Medical reasons involve the woman’s endangered physical and/or mental health, a risk to the woman’s life, and incurable malformations or illness in the foetus.M.T. Munoz Sastre, C. Petitfils, P.C. Sorum & E. Mullet, ‘A mapping of the positions of adults in Toulouse, France, regarding induced abortion’, European Journal of Contraception and Reproductive Health Care 20(2015): 158-169.

Our hypothesis was that the five personal positions reported by Muñoz Sastre et al. would be found among participants of each religious group, although the percentage of participants sharing each of these positions would strongly differ from one religious group to another. These hypotheses were grounded on the fact that (a) as shown in previous studies, French Catholics are deeply divided on the issue of abortion and few of them follow their leaders; (b) as suggested by several authors, dissentions on this issue may also arise among Evangelicals, although to a much lower extent than Catholics;B.G. Smith & B. Johnson, B., ‘The liberalization of young Evangelicals: A research note’, Journal for the Scientific Study of Religion 49(2010): 351-360. J. Farrell, J., ‘The young and the restless? The liberalization of young Evangelicals’, Journal for the Scientific Study of Religion 50(2011): 517-532. and (c) the same differences in opinion seem to affect non-believers as well and for reasons completely separate from religion.T. Palermo, M.I. Erazo, V. Hurtado Pinochet, ‘Women’s opinions on the legalisation of abortion in Chile 2009-2013’, Culture, Health & Sexuality 17(2015): 873-890. J. Valencia Rodríguez, K.S. Wilson, C. Díaz Olavarrieta, S.G. García & M.L. Sánchez Fuentes, ‘Public opinion on abortion in eight Mexican states amid opposition to legalization’, Studies in Family Planning 42(2011): 191-198.

Method

Participants

The participants were unpaid volunteers recruited and tested by one of the authors (CM) and several research assistants. They first contacted members of Evangelical churches, explained the study, asked them to participate, and, if they agreed, arranged where and when to carry out the experiment. Of 65 people contacted, 43 (66%) participated, 30 women and 13 men. The mean age was 28 years; 41 (95%) of them considered themselves as regular attendees. The experimenters contacted people in the street and tried to constitute two other samples – one sample of Catholics and one sample of non-believers – which were equivalent to the first one in terms of age and gender composition. Of the 95 people contacted who identified themselves as Catholics, 48 (51%) were included in the study, 33 women and 15 men. Their mean age was 31 years; 18 (37%) of them considered themselves as regular attendees. Of the 90 people that declared they do not believe in any God, 45 (51%) were included in the study, 31 women and 14 men. Their mean age was 31.   

Material

The material, taken from Munoz Sastre et al., consisted of 45 cards containing a story of a few lines, a question, and a response scale. The stories were composed according to a three within-subject orthogonal factor design: (a) the reason for abortion (personal convenience, poor economic situation, pregnancy from rape, foetal malformations or the mother’s life at risk with pregnancy), (b) the gestational age (about six, four and a half, or two months), and (c) the person’s age (about 42, 30, or 18 years). In these stories, the male partner always agreed with the woman’s request for an abortion. There were 45 different stories.

An example of a story is the following: ‘Mrs. Henry is 41 years old. She is about fourth and half months pregnant. Mrs. Henry wants to get an abortion because she was told by her obstetrician that the foetus has congenital malformations. The obstetrician has been very clear on this point. Her partner agrees with her request.’ Along with each story there were a question and a response scale. The question was ‘To what extent do you believe that Mrs. Henry’s abortion would be an acceptable medical/surgical procedure in this case?’. There was an 11-point scale where 0 at the left end of the response scale meant ‘Not at all acceptable’ and 10 at the right end of the response scale meant ‘Completely acceptable’. Cards with the stories were arranged randomly for each participant. All the participants answered additional questions about their age, gender, educational level, and religious involvement.

Procedure

The site was either a vacant classroom in the local university or the participant’s private home. Each person was tested individually. The procedure followed Anderson’s recommendations for this kind of study.N.H. Anderson, ‘Unified social cognition’, New York, NY: Psychology Press (2008). N.H. Anderson, ‘Unified psychology based on three laws of information integration’, Review of General Psychology 17(2013): 125-132. The participants took 30-45 minutes to complete the ratings. The study complied with the ethical recommendations of the French Society of Psychology, i.e. full anonymity was respected and informed consent was obtained from all participants in the study.

Figure 1. The five clusters. Mean acceptability ratings are on the vertical axis. Reasons for abortion are on the horizontal axis. Each curve corresponds to one gestational age. Each panel corresponds to a cluster. E = Evangelicals, C = Catholics, A = Atheists.

Results

A cluster analysis was performed on the data.J. Hofmans & E. Mullet, ‘Towards unveiling individual differences in different stages of information processing: A clustering-based approach’, Quality and Quantity 47(2013), 555-564. As in Munoz Sastre et al., a five-cluster solution was retained. The data patterns relevant to the clusters are shown in Figure 1 and the distribution of participants in each cluster is also shown. Separate ANOVAS were conducted on the data of each cluster, using a Reason (5) x Gestational age (3) x Mother’s age (3) design. Detailed results are available from the corresponding author. A rating of attendance to church was computed and ranged from 0 (no attendance) to 2 (regular attendance).

Never acceptable cluster

For 30 participants (22%), acceptability ratings were always close to the left end of the response scale (M = 1.22).  This cluster was the expected Never acceptable cluster. Mean acceptability ratings were slightly higher for endangered health (2.32), rape (1.51) and foetal malformation (1.50) than personal convenience (0.18) or socio-economic problems (0.58). 97% of the members of this cluster were Evangelicals and the remaining 3% were Catholics. The mean attendance rating of this cluster was 2.00; that is, everyone in this group was a regular attendee.

Reason cluster

For 27 participants (20%), the reason given by the mother was by far the most important factor. Mean acceptability ratings were very high for endangered health (8.64), rape (8.10), and foetal malformation (7.97), and very low for personal convenience (1.10) and of socio-economic problems (2.00). This cluster was the expected Reason cluster. 22% of the members of this cluster were Evangelicals, 45% were Catholics, and the remaining 33% were non-believers. The mean attendance rating of this cluster was 1.07.  

Reason and gestational age cluster

For 25 participants (18%), both the reason given by the mother and the gestational age were important factors.  Mean acceptability ratings were high for endangered health (7.68), rape (7.56) and foetal malformation (7.23), medium for socio-economic problems (6.03), and low for personal convenience (2.16). Mean ratings were higher when gestational age was two months (6.76) than when it was six months (5.51). Also, they were slightly higher when the mother was 42 (6.32) than when she was 30 (5.94). This cluster was the expected Reason and gestational age cluster. 28% of the members of this cluster were Evangelicals, 32% were Catholics, and the remaining 40% were non-believers. The mean attendance rating of this cluster was 0.96.  

Legalist cluster

For 25 participants (18%), both the reason given by the mother and the gestational age were important factors, and more importantly, they interacted. When the gestational age was two months, ratings were always high; they ranged from 7.99 (personal convenience) to 9.81 (rape). When the gestational age was six months, ratings strongly varied as a function of reason. They varied from 1.03 (personal convenience) to 8.64 (endangered health). This cluster was the expected Legalist cluster. 52% of the members of this cluster were Catholics, and the remaining 48% were non-believers. The mean attendance rating of this cluster was 0.64.  

Always acceptable cluster

Finally, for 29 participants (21%), acceptability ratings were always relatively high. This cluster was the expected Always acceptable cluster. Mean ratings were very high for rape (9.71), endangered health (9.61), and foetal malformation (9.61), and high for socio-economic problems (7.58). Mean ratings were lower in case of personal convenience but still higher than the middle of the response scale (5.76). 2% of the members of this cluster were Evangelicals, 48% were Catholics, and the remaining 48% were non-believers. The mean attendance rating of this cluster was 0.72.  

 The five clusters were significantly different in terms of religious affiliation, Chi²(8) = 82.94, p < .001, and in terms of religious involvement (attendance rating), F(4, 131) = 16.38, p < .001.

Discussion

Among Evangelicals, three main positions were found: roughly two-thirds strictly aligned with the position of their leaders and denied that having an abortion can sometimes be an acceptable decision, whereas (a) one in six considered that, in case of rape, malformation and endangered mother’s life, abortion was acceptable, and (b) one in six considered that in the three preceding cases plus the case of poverty, abortion was acceptable and still more acceptable if the foetus was 2 months old rather than 6 months old. There was even one participant who expressed the view that the decision to have an abortion was acceptable in most cases. This finding was consistent with suggestions by Farrell and by Smith and Johnson that even among Evangelicals there is a diversity of opinions on the decision to have an abortion.

Among Catholics, four main positions were found, each one being endorsed by roughly one quarter of them. Only one Catholic participant strictly aligned with the position of the Holy See, and denied acceptability to induced abortion in all cases. The main opposition in this group was between Catholics who disagreed with the first part of the French law by denying acceptability to abortion for personal convenience when the foetus was 2 months old, and those who disagreed with the law by considering that abortion for personal convenience was acceptable even if the foetus was older than 2 months. Catholics who expressed a ‘legalist’ view were, as a result, a minority. This finding was consistent with previous findings showing, at least in some countries, a diversity of opinion among Catholics and a general lack of alignment with the Holy See’s positions. Future studies should examine more in depth the reasons why some people identify themselves with the catholic faith, practise their faith on a regular basis, but believe that abortion can be acceptable even if the foetus is four to six months old.

Among atheists, four positions were also found, and the proportions of participants sharing each of these views were not very different from the ones observed in the group of Catholics. Studies showing a similarity of views on abortion between Catholics and Atheists are rare but such a result has been found by Valencia Rodríguez et al. who concluded that contrary ‘to the findings of previous public opinion studies in Mexico … sex and frequency of religious service attendance were not significant predictors of abortion opinion’ (p. 196). Future studies should examine, more in depth, the reasons why a non-negligible proportion of people who do not identify themselves with any faith are, nevertheless, of the opinion that abortion is never acceptable without ‘solid’ reasons such as foetal malformation or endangered mother’s health. During data gathering some people from this group expressed the view that aborting is killing, that killing without a good reason (e.g., self-defenceD. Hermand, E. Mullet, P. Tomera & V. Touzart, ‘The relationship between intent, consequences, the dangerousness of the victim, and blame: The case of self-defense’, Psychology, Crime and Law 7 (2001), 57-69.) is, generally speaking, a crime.

Unsurprisingly, religious involvement was associated with the five personal positions. Three levels can be distinguished. High religious involvement (church attendance) was associated with the ‘never acceptable’ position. Medium religious involvement (belief in God) was associated with the two ‘reason’ positions. Absence of religious involvement was associated with the ‘legalist’ and the ‘always acceptable’ position.

Limitations

The group of participants did not constitute a representative sample. The objective of the study was not to assess the specific percentage of the general population who would hold each of the five positions – it was in no way an epidemiological study – but to examine the effect of religious affiliation on these positions. As a result, three groups roughly equivalent in terms of number, age and gender, were constituted and compared. Generalisations to other religious affiliations or to other countries must be made with care.

Implications

This study found a remarkable consensus regarding abortion between those who identify themselves as Catholics and those who call themselves atheists. This reflects in part that most nominally Catholic French people may not truly believe in all Catholic doctrines. But this consensus – as well as French law – is challenged by the rise of Evangelicals who are quite adamantly opposed to abortion, mostly under any conditions. Therefore, the Evangelicals are providing support to the Catholic Church’s hierarchy in this issue. With the rise of Evangelical churches, the battle over abortion is, as a result, likely to become more heated.