Suicide beliefs of Jews and Protestants

In Jewish tradition, the “duty of preserving life, including ones own, is considered paramount” (1). Although suicide in times of religious persecution is regarded as meritorious (Kiddush HaShem: lit “Sanctification of the name (of G-d)”:dying to prevent transgression), suicide is otherwise viewed as “a most heinous sin, even worse than murder” (1). In Protestant Christianity, suicide has in the past been regarded as a serious sin, and a criminal act. For example in the UK, attempted suicide could be subject to criminal prosecution, and in deed was, until well into the twentieth century. However in many contemporary predominantly Christian societies, including the UK, social and legal attitudes towards society have become more tolerant and condoning throughout the last century (2).

In this study, we looked at the relationship between cultural-religious factors – Jewish vs Protestant background - and beliefs about suicide, and reported suicidal behaviour. There have been cultural-religious differences in suicide rates, and we ask whether these are related to differences in individual beliefs.

Early studies found rates of suicide to be low within Jewish communities. For example, in a 1960s report, suicide in the Jewish population was said to be rare (3). Suicide rates were noted to be low in Israel (4). More recently, it has been noted that suicide rates for Israel are generally lower than those of other countries, including America (5). Kohn et al (6) concluded that ‘the suicide rates amongst the youth in Israel, as in adults, are amongst the lowest in the world’. Although there was evidence, by the mid-1990s, that suicide rates among young adult male Jewish Israelis had climbed (7), this trend had not continued into the late 1990s (8), and overall Jewish suicide rates in Israel continued to be low compared to those in other countries.  Comparative studies have found suicide rates to be lower amongst Jews than the general population in predominantly Protestant communities (9). For example, an eleven-year study looking at rates of suicide in Michigan indicated the rate of non-Jewish suicide to be 14.1 per 100,000, but only 6.9 per 100,000 among Jews (10).  2.9% of a US adult sample had attempted suicide at least once in their lifetime (2), whereas this figure was only 1.4% in a Jewish sample (11). The latter sample also reported lower rates of suicidal ideation than do adults in predominantly Protestant countries.

Much less research has been carried out examining cultural-religious differences in beliefs about suicide as opposed to actual suicide. A number of studies have, however, compared Christian beliefs with those of other religious groups in which suicide is not condoned. For example, a primarily Moslem group of Nigerian students saw suicide as less acceptable and normal than did a mainly Christian group of Zambian students (12). Nigerian students also viewed suicide more negatively than did American students (13). In the USA, views towards suicide in special circumstances are often liberal. Thus 65% of medical students favoured the legalization of  Physician Assisted Suicide (PAS) (14).  Among emergency medical technicians, 68% agreed that PAS should be legal, and 77% agreed that terminally ill people have the right to commit suicide (15). None of these studies included Jews, so we do not know whether Jewish beliefs about suicide reflect the reported low suicide rates among Jews. 

In this study we tried to fill this gap in knowledge. As well as looking at Jewish-Protestant differences in beliefs about suicide, we controlled for the effects of factors which might impact on beliefs about suicide, and which might be confounded with the effects we were interested in: age, religiosity, depression, and anxiety. We also examined gender differences in beliefs about suicide, since it was thought that these might interact with cultural-religious effects. Among Jews, gender differences in suicide rates have been reported as not marked (4), whereas in Protestant groups, men have a higher rate of completed suicide than do women, but lower rates of parasuicide (2). 

Method

Design

The study sought to examine the effects of cultural-religious group (Jewish versus Protestant), and gender, on beliefs and behaviour regarding suicide. Potential extraneous variables (age, religiosity and levels of depression and anxiety) were statistically controlled through the use of analysis of covariance. Sample size for the four sub-groups is based on a power analysis (16), giving a minimum of 35 per group as sufficient to detect a medium effect, with alpha <.05, and a statistical power specification of 0.80.

Participants

A quota sampling strategy was employed. We aimed to recruit approximately equal numbers of Jews and Protestants, men and women, while attempting to secure a range of religious practice and affiliation (including “nones”),  and groups that were comparable in terms of age, occupational status and other factors. Jewish participants were recruited via synagogue membership lists and by snowballing among the non-affiliated. Protestant participants were recruited via church membership lists and by snowballing among the non-affiliated. Such methods have been used in research of a similar nature (17, 18).  This method of recruitment secured participants with a wide range of religious practice and affiliation, with a Protestant or Jewish family of origin.  

There were 161 participants, 70 Jewish (35 men, 35 women) and 91 Protestant (44 men, 47 women). The mean age of all participants was 40.78 years (S.D.=15.75 years). 64% were in steady relationships (married, engaged or cohabiting), while 36%  were single, divorced, separated or widowed. 75% were graduates and/or employed on a professional basis, and 25% were in other occupations, including white- and blue-collar workers, homemakers and retired people.  The four comparison groups, Jewish Men (JM), Jewish women (JW), Protestant Men (PM) and Protestant women (PM) were comparable in terms of age, marital status, occupational status, and the potential confounding variables that were in any case partialled out in the analysis (religious activity, depression, anxiety). 

Materials

The reasons for living  (RFL) inventory. (19). This self-report instrument measures a range of beliefs that may be important in not committing suicide. The questionnaire consists of 47 reasons, grouped into six scales. Participants indicate how important each reason for living would be to them if suicide were contemplated, on a six point scale. There are six scales: Survival and coping beliefs e.g., ‘I would see no reason to hurry death along’; Responsibility to family  e.g. ‘I love and enjoy my family too much and could not leave them’ ; Child-related concerns e.g. ‘I want to watch my children as they grow’; Fear of the actual act of suicide e.g. ‘I could not decide when, where and how to do it’; Fear of social disapproval e.g. ‘I would be concerned about what others would think of me’;  Moral objections e.g. ‘I believe that only God has the right to end a life’ (In this study, one item, “I would be afraid of going to hell”, was removed, as the concept of hell was thought to differ between groups). A higher score indicates greater agreement with the reasons not to commit suicide. Cronbach’s alpha for survival and coping beliefs = 0.96; for family beliefs = 0.81; for children beliefs = 0.65; for fear of suicide beliefs = 0.87; for social disapproval beliefs = 0.85; for morality beliefs (modified by the removal of the item about fear of hell) = 0.94. This measure has been highly recommended as a suicide assessment instrument (20), and has demonstrated good psychometric properties, including sufficient construct validity (21) and sensitivity in detecting ethnic differences in reasons for living (22).

Acceptability of suicide in specific situations. Participants were provided with five different situations (e.g. ‘has an incurable disease’ and ‘has gone bankrupt’), and were required to indicate whether somebody would have the right to end his/her life in each situation, with yes, no or don’t know. 

Suicidal ideation. Suicidal ideation was measured using five relevant self-report items from the Present State Examination (23). e.g., ‘Have you ever wished that you were no longer alive?’ and ‘Are you ever afraid that you will harm yourself ?’). Reliability of the scale within the sample was good (Cronbach’s alpha = 0.85). 

Attempted suicide. Using self-report, participants were asked to simply indicate whether they had ever actually tried to kill themselves by circling one of two responses; yes or no.

Several factors were assessed and used as covariates in the statistical analysis. As well as age, these were: 

Religious activity (24). Three kinds of religious activity common to both religious traditions were rated for frequency on a five-point scale: attending a church or synagogue, praying, and studying religious texts. The scale has been successfully used in several studies involving several religious groups (24) Reliability in the current study was good (Cronbach’s alpha = 0.91).

Depression and anxiety (Hospital Anxiety & Depression Schedule:HADS) (25): This self-report questionnaire consists of seven depression items (e.g. ‘I still enjoy the things that I used to enjoy’) and seven anxiety items (e.g. ‘I get sudden feelings of panic’). Cronbach’s alpha for the depression subscale = 0.77; Cronbach’s alpha for anxiety subscale = 0.82.

Procedure

Participants were asked to complete the questionnaires in their own time, and were told that they were free to withdraw or omit any items, and that their answers would be anonymous. Questionnaire booklets were returned in pre-paid addressed envelopes. A total of 271 questionnaires were sent or given out, of which 170 were returned (a response rate of 63%). Nine of these were excluded from analysis, because the current religious affiliation of participants did not match their religious background. The final sample therefore comprised 161 participants.

Results

Rates of attempted suicide. Of the 63 Jewish respondents who answered this question, two had previously attempted suicide (3.1%). Of the 85 Protestant respondents who answered this question, two had previously attempted suicide (2.4%). Further analyses were not carried out on these data in view of these low frequencies and negligible group differences.

Suicidal Ideation (SI).The effect of religious affiliation and gender was examined with a 2 (Jewish v Protestant) x 2 (Male v Female) ANCOVA, with age, depression, anxiety and religiosity entered as co-variates.  Scores on a 0-10 scale were:

Jewish men: 2.94 (sd 3.34);

Jewish women: 2.58 (sd 3.28);

Protestant men: 2.24 (sd 2.94);

Protestant women: 3.16 (sd 3.35);

There were no significant effects of cultural-religious group or gender, or their interaction. 

Reasons for living.Reasons for living scores can be seen in Table 1. The effect of religious affiliation and gender was examined with a 2 (Jewish v Protestant) x 2 (Male v Female) ANCOVA, with age, depression, anxiety and religiosity entered as co-variates. 

INSERT TABLE 1 ABOUT HERE

It can be seen from Table 1 that cultural-religious group had a marginal effect on fear of suicide and moral objections, with Jewish participants displaying stronger beliefs that these were reasons for living. Women showed more fear of suicide than men. In addition, there was a significant interaction between religious affiliation and gender on survival and coping beliefs. Post-hoc contrasts (Scheffe) indicated that within the Jewish participants there were no gender differences (p>.05), while among the Protestants, men had significantly higher belief scores than women (p<.05). There were no effects of religion or gender on responsibility to family or child related concerns. It should be noted that on both  these scales, the scores were rather high, almost approaching ceiling level. 

Acceptability of suicide according to circumstances.Participants were given a number of scenarios and for each, were asked whether one would have a right to end their life faced with that situation. 

INSERT TABLE 2 ABOUT HERE

Jewish and Protestant participants were marginally significantly different from each other with respect to how they responded to items 1 and 4: Jews were less likely to feel that an individual had the right to end their life. There were no religious group differences for the other items. A comparison of responses by gender revealed that overall, men and women did not differ in their responses. However, there were some group x gender interactions. For example, for item 2 (…has gone bankrupt), Protestant men were more likely to think that suicide was acceptable in this situation than Protestant women (X² (2, n=89) = 6.73) p<.05).  This was also the case for item 3 (…has dishonoured his/her family): X² (2, n=88) = 6.54, p<.05). There were no differences between Jewish men and women in their beliefs about the acceptability of suicide.

Relations of depression, anxiety, and religious practice with suicide ideationand beliefs.

Finally it is noted that although the four groups were comparable in terms of  the potential confounded variables, we did note that religious practice, depression and anxiety all related in an expected direction, and often signifcantly, with many of the dependent variables (suicide ideation and beliefs) that we were interested in. To summarise, depression and anxiety went with higher suicide ideation(r=.25, p<.001, and .21, p<.01, respectively),  with beliefs in the right to kill oneself under certain circumstances (with total beliefs in the right to kill oneself, r=.12, ns, and .07, ns, respectively), and with weaker endorsement of some reasons for living (with total RFL, r=-.32, p<.001, and -.11, ns, respectively), while higher religious activity showed an opposite pattern of correlations – going with lower suicide ideation (-.10, not significant), lower beliefs in the right to kill oneself under certain circumstances (with total beliefs in the right to kill oneself, r=-.52, p<.001), and stronger endorsement of some reasons for living (with total RFL, r=.29, p<.001). This confirms the validity of the measures we were using.  However, more detailed statistics will not be presented, since the effects of depression, anxiety and religiosity on suicide ideation, behaviour and beliefs were not the main focus of the study, and were partialled out by analysis of covariance.

Discussion

In this study we asked whether Jews and Protestants differed in their beliefs about suicide, and whether these differences might relate to differences in the prevalence of suicide in these groups.

This study failed to show differences between the two groups of participants in actual suicide behaviour and ideation. Sample size was small, but there was no evidence of even a small difference in the expected direction. Moreover sample size had been calculated to detect likely effects, and many other effects were indeed detected.

Jews and Protestants did report some differences in suicide-related beliefs, controlling for the effects of variations in levels of religious activity in our analyses. The differences were not apparent on every one of the measures we used. Jews were more likely than Protestants to report moral/religious reasons for living, and also to report fear of the act of suicide itself. Jews were more likely than Protestant women (but not men) to endorse survival and coping beliefs, and Jews were marginally less likely than Protestants to agree that one has the right to end life if one has an incurable disease, or if one is tired of living. The package of beliefs on which we observed cultural-religious group differences is, on the whole, consistent with the view that Jewish culture does not condone suicide to the extent that Protestant culture now does. 

But there are several points that call for comment.

First, the fear of suicide, reported more strongly by Jews than by Protestants, could have only a very tenuous connection with the culturally-carried package of beliefs condemning suicide.

Second, there is a well-documented Jewish tradition of approval for suicide in conditions of religious persecution (1) – for example the mass suicide of large numbers of Jews at Masada during the Roman occupation of Israel – and this study did not examine beliefs about this type of suicide. We cannot say whether this is currently seen as acceptable among Jews.

Third, we observed no Jewish-Protestant differences in beliefs about responsibility to family, and child-related concerns, as reasons for living. Judaism is often said to be a family-centred religion, with a strong emphasis on the value of child-rearing, and of religious activities in the home and family setting (26). We might expect that child and family reasons for living might be more strongly endorsed by Jews than by Protestants, but there were no group differences. However we noticed that these reasons for living were strongly endorsed by both groups, and there was a near ceiling effect. Thus family-related reasons for living were high among Jews, and also among Protestants.

Finally, we noted that gender differences in suicide beliefs were less marked among Jews than among Protestants.  Among Protestants, men were more likely to say that suicide was acceptable under some circumstances, than were women, and they were less likely to endorse fear of suicide as a reason for living. These are consistent with reported gender differences among Protestants in suicide rates. However Protestant women’s lower survival and coping beliefs do not readily fit with this. Protestant women’s higher rates of parasuicide compared to Protestant men are perhaps consistent with their lower coping and survival beliefs, and their higher fear of the act of suicide, but this is very post-hoc. It is very difficult to propose a compelling match between gender variations in beliefs about suicide, and the fact the (Protestant) men are less likely than women to attempt suicide, but more likely to complete it. However we did note that gender differences among Jews with respect to suicide beliefs were not marked, and this is consistent with some observations that gender differences in suicide rates among Jews have not always been marked.

In conclusion, we think that the Jewish-Protestant differences in suicide beliefs observed in this study are broadly consistent with what has been reported about Jewish-Protestant differences in suicide prevalence. We cannot make generalisations about Jews and Protestants in general based on this small sample of UK residents. In particular we cannot be sure whether similar findings would be made among young adults in contemporary Israel, where young men may be more likely to commit suicide than they have been in the past. We also cannot infer much about causal relations, but it is tempting to speculate that culturally-carried beliefs about suicide may have an impact on suicide behaviour. 

Acknowledgements

The research reported in this article was supported with funding from the Economic and Social Research Council (Research Grant No. R000 22 2685). We would like to thank Professor Robert Kohn for very generously sharing his collection of material on Jewish suicide.