This publication is copyright J.J. Sexton and Richard O'Leary (1996) and is included on the CAIN site by permission of Blackstaff Press and the authors. You may not edit, adapt, or redistribute changed versions of this for other than your personal use without express written permission. Redistribution for commercial purposes is not permitted.

The papers contained in this book were commissioned by the Forum for Peace and Reconciliation. The views expressed are the views of the authors and are not necessarily those of the Forum or its members.

CONTENTS

PREFACE BY JUDGE CATHERINE McGUINNESS,
CHAIRPERSON OF THE FORUM FOR PEACE AND RECONCILIATION

1. Obstacles to Reconciliation in the South
BY ARTHUR AUGHEY

2. A Unionist Legal Perspective on Obstacles in the South to Better Relations with the North
BY BRICE DICKSON

3. The Role of the Catholic Church in the Republic of Ireland 1922 - 1995
BY DERMOT KEOGH

4. Religious Minonties in the Irish Free State and the Republic of Ireland (1922 - 1995)
BY TERENCE BROWN

5. Factors affecting Population Decline in Minority Religious Communities in the Republic of Ireland
BY J.J. SEXTON AND RICHARD O’LEARY

Notes on Contributors

FACTORS AFFECTING
POPULATION DECLINE IN
MINORITY RELIGIOUS COMMUNITIES
IN THE REPUBLIC OF IRELAND

J.J. SEXTON
AND
RICHARD O'LEARY

ACKNOWLEDGEMENT

The authors wish to express their appreciation of the assistance received from the staff of the Central Statistics Office, Dublin (particularly Mr Aidan Punch and Mr Gerard Keogh) for the supply of special tabulations from past censuses of population.

CONTENTS

SUMMARY

1. INTRODUCTION

2 Data sources: classifications
Sources
Classifications

3 General trends
National trends
Geographical aspects
Age structures

4. Factors affecting the trend in the numbers in the minority religious communities
Fertility
Nuptiality
Mixed marriages

5. Birth and death rates for different religious groups Rates of natural increase
Birth rates
Death rates
Rates of natural increase

6. Migration
Some methodological aspects
Migration from 1911 - 1971
Estimating more recent trends 1971 -1991

7. The Jewish community
General trends
Age Structure
Rates of natural increase
Migration
Mixed marriages

NOTES

REFERENCES

APPENDIX TABLES

APPENDIX I Census data on fertility of marriage

APPENDIX II The compilation of death rates for different religious groups

4
FACTORS AFFECTING THE TREND
IN THE NUMBERS
IN THE MINORITY RELIGIOUS COMMUNITIES

The significant changes which have affected the minority religious communities in Ireland illustrated in the preceding section immediately prompt questions as to why these changes occurred. This section of the report is concerned primarily with a number of such causative issues. It is important to bear in mind however, that causations and outcomes can be interactive in the particular context involved here, especially over time. It is clear, for example, that various influences (e.g. migration) must have played a role in bringing about the changes in the age structure of the minority religious communities. However. these changes can, in turn, exert further causation. Changing age structures can, for example, exert an influence through their effects on birth and mortality levels and, as a consequence, on the rate of natural increase in a population (or subpopulation). However, the primary causative factors which we will discuss in this section relate to fertility, nuptiality, and the related question of mixed marriages. Each of these three influences will now be considered in turn. Migration is clearly a further candidate for consideration in this regard, but we will defer discussion of this aspect until section 6 as it fits more logically into the sequence of our analysis at that point.

FERTILITY

The usual method employed to analyse fertility within a population is to relate numbers of births in a given period (usually a year) to the female population in the fertile age groups (i.e. between 15 and 44 years). However this cannot be readily done in the context of comparing religious denominations as the official birth statistics do not involve a classification by religion. One can, however, by using census data classified by religious group, estimate relevant rites by relating the numbers of very young children enumerated (e.g. aged less than 1 year, less than 2 years, etc.) to the number of women in the above - mentioned fertile age band. Such estimations would not, of course, yield the same results as those based on birth registration data (it would, for example, exclude children who died soon after birth) but they would not differ significantly and they should be sufficient to show up any relative differences between religious groups, especially if these are substantial. Such calculations are given in Table 4 for marital births only in the form of estimated annual average births per 1,000 married women aged 15 to 44 years. The analysis is restricted to births within marriage as those still form the great majority of all births and there is still a much greater propensity for married women to have children, even though the proportion of extra-marital births in the overall total has been increasing significantly in recent years.⁹ This approach is basically the same as that followed by Walsh (1970), in fact the figures represent a continuation of his data which related to 1946 and 1961.¹⁰ In addition to the actual indicators as described, Table 4 also contains ratios (in column 4) which show the relationship between the fertility rates for Catholics and the minority communities.

Sources:
(1) Walsh, B.M., Religion and Demographic Behaviour in Ireland, ESRI, 1970 General Research Series. Paper No 55.
(2) Censuses of population 1971, 1981, 1991.
Notes:
(1) The annual average number of births in each year was calculated by taking the population aged 0 to 2 and dividing by 3. The number of marital births for each religious group was then estimated by applying the national proportion of such births.
(2) For the years 1946 and 1961 the category ‘other denominations’ includes those who indicated that they had ‘no religion’ and those who did not provide information in the censuses on religious affiliation. For these years, however, the numbers involved for these two groups were extremely small. For later years (1971, 1981, 1991) the ‘other denominations’ category consists of the main Protestant religions and ‘other stated religions’, but excludes those with no religion and persons who declined to provide information on their religion in censuses. It is considered that this procedure ensures that the fertility indicator for the ‘other denominations’ group is reasonably consistent over the entire period in question.

The figures show that Catholic fertility levels have been consistently higher than those for the minority communities. Between 1946 and 1981 the relative excess was, broadly speaking, more than 55 per cent. During the 1980s, however, it is clear that Catholic fertility fell much more rapidly with the result that by 1991 the relevant indicator for this group was some 30 per cent higher than that for the minority group (the actual figures being 118.5 and 91.4). In demographic terms this (1991 relative excess, though less than for earlier periods, is still of significant proportions.

The relative differences in fertility levels between Catholics and the minority community as indicated by this analysis are not unexpected. As has long been acknowledged, and demonstrated, there are significant differences in fertility levels between the Catholic and minority communities population.¹¹ However, a somewhat surprising feature of the results is the extent of the fall in fertility for the minority communities over the period concerned, which was not a great deal less than that for the Catholic community. Between 1946 and 1991 Catholic fertility as measured by the indicators in Table 4 declined by over 55 per cent while that for the minority religious communities group decreased by almost 50 per cent. The decrease in fertility for the latter was particularly rapid during the 1970s when the decline equalled that for Catholics. One would, perhaps, have expected smaller decreases for the minority community, given the much lower actual fertility levels which initially prevailed.

There are possible reasons for this which are not necessarily related to fertility as such. One relates to the incidence of mixed marriages which, as we shall subsequently illustrate, can exert a significant influence on the indicators involved here. This is because of the likelihood of the relative displacement of mothers and children in a mixed marriage situation in regard to their inclusion (or exclusion) in the numerator and the denominator in the above mentioned index. If, for example, a mother from the minority communities has children who are Catholics (and this, as we shall later demonstrate, is the most likely outcome in a mixed marriage) she is included in the denominator of the relevant index for that community but her children are not. This has the effect of reducing the value of the index for the minority group, and increasing it for Catholics, even though the impact in the latter case is negligible due to the overall size of the Catholic population. This issue is discussed in more detail later in this section of the report.

There is an interesting corollary to this. If one accepts the proposition that mixed marriages were a significant factor in curtailing the number of births in the minority communities (and later evidence presented suggests that this was indeed the ease) then it is likely that actual or real fertility levels within these communities may not have fallen all that much in the post war period; in fact they may have even increased slightly.

There are, of course, other factors which can exert an influence on the differences in the fertility levels indicated for the two groups in question. One such influence is social group structure which is significantly different for Catholics and the minority religious communities. This aspect is considered in sonic detail Appendix I. This, however, shows that variations in social structure between the two denominations would not, of themselves, give rise to the differences between the aggregate fertility levels set out in Table 4. Differences in the age distribution of married women in the fertile age band (15 to 44 years) for the two religious groups is another such possible influence. However, an inspection of these distributions for the two denominational categories for 1946 and 1991 reveals little difference between them, in either year. Furthermore, these structures have changed little over time, and to the extent that they have (e.g. relatively larger proportions in the younger age groups) they have applied equally to both groups. Thus, this factor is unlikely to have affected the fertility comparisons given in Table 4, either in cross-sectional terms for any one year, or over time.¹²

NUPTIALITY

Let us now consider the question of marriage rates or the degree of nuptiality in the population as a whole for different religious groups. This is still an important aspect to consider, even when viewed against the background of increasing numbers of extra-marital births, as married women are still much more likely to bear children than single women. Births within marriage still account for over 80 per cent of all births. It is, of course, even more important in analysing trends for past periods when the proportion of extra-marital births was very much smaller.

Sources:
(1) Walsh, B.M., (1970), Religion and Demographic Behaviour in Ireland, ESRI, General Research Series. Paper No 55.
(2) Censuses of population.
Notes:
(1) The nuptiality rates are the numbers of married women aged 15 to 44 vein expressed as a proportion of all women in this age group.

As in the case with the fertility related data, this analysis has to be confined to census years in which questions on religious affiliations were asked. In this (census) context the indicator used is a simple one, namely the number of married women in the fertile age group, 15 - 44 years, taken as a percentage of all women in this age band. The relevant figures are given in Table 5 covering census years from 1946 to 1991.¹³

This indicator shows that Catholic nuptiality has been consistently lower than that for the minority communities, broadly speaking by about 10 per cent. The evidence indicates that the divergence was somewhat greater in earlier periods (e.g. in the 1940s). The nuptiality rates increased for both communities between 1946 and 1981; both rates then declined markedly in the 1980s, even though the relationship between the two indicators remained largely unchanged.

Walsh ( 1970) noted that nuptiality for the minority communities in the Republic, while higher than that for Catholics, was lower than that for the Protestant community in Northern Ireland. He suggested that this may have been due to the small size of the minority religious groups in the Republic, combined with a desire to avoid intermarriage.

The nuptiality differentials as indicated serve to increase the number of births in the other denominations population relative to those for Catholics. However, it is clear that the scale of the divergence is not substantial and is not sufficient to offset the effects of the relatively lower fertility associated with the minority religious group.

MIXED MARRIAGES

While the somewhat higher incidence of marriage for the minority communities population contributes to increasing the numbers of births for that group relative to the Catholic population, there is another marriage related feature which tends to offset this. This concerns the question of inter-Church or mixed marriages, i.e. where parents are of different religious affiliations. In this context most interest centres on situations where one of the parents is Catholic in view of the application by the Catholic Church of the Ne Temere decree, particularly its provisions regarding the upbringing of children.¹⁴ The following analysis, therefore, relates exclusively to circumstances where one parent is Catholic and the other is of another specific denomination, without any attempt being made to further subdivide the latter category. We have already commented, in the earlier section on fertility, how mixed marriages can influence the indicators of relative fertility as between the denominational groups in question.

In order to provide some information on the exact circumstances which prevail in regard to mixed marriages special tabulations were extracted from the 1981 and 1991 censuses which identified families living in private households in which parents were of different religions and which provided counts of the children (of any age) in these families, also classified by religion.¹⁵ The results are summarized in Table 6. This shows that in 1981 there were 6,570 such families enumerated with 14,813 children, of which 12,670 (over 86 per cent ) were Catholic. There is some evidence that the Catholic proportion of children tended to be somewhat lower for mothers in the younger age groups. In the 1991 census the number of mixed marriage families had increased to 9,110 and the number of children to 20,420; on this occasion the number of Catholic children, though still substantial at 15,886, constituted a somewhat smaller proportion of the total, nearly 78 per cent.¹⁶

The data (not shown in the table) also distinguished the religion of the parents and from these there is some evidence that the Catholic proportion of children in a mixed marriage tends to be somewhat higher when the mother is Catholic. The relevant proportions of Catholic children born to Catholic and minority group mothers in 1991 were 81 and 73 per cent respectively.

Sources:
Special tabulations from the 1981 and 1991 censuses.
Notes:
(1) Couples without children are not included.
(2) The children included in this table are of any age.

Source:
Special tabulations from the 1981 and 1991 censuses.

While these figures (not previously compiled) are of great interest, their real significance emerges when they are compared with comparable data for homogamous Catholic or minority communities families (i.e. in which both parents are either Catholic or of another religious group). This further information is given in Table 7 and shows that in 1991 the number of mixed marriage families (some 9,100) was quite large when viewed in the context of the number of homogamous minority communities families (14,500); the corresponding numbers of children were 20,400 in mixed marriages and 32,400 in homogamous minority communities families.

There are other ways of viewing this aspect, for example by considering rates based on the numbers of married persons from the minority communities who were in mixed marriages taken as a proportion of all married persons from the same religious group. This calculation yields the following results for 1981 and 1991.

These figures show that in 1991 nearly one quarter of all married persons from the minority religions communities were in mixed marriages. The proportion for males, over 26 per cent, was noticeably higher than for females, 21 per cent. This proportion has been increasing in recent decades; the corresponding overall percentage for 1981 was significantly lower, less than 19 per cent. In fact these data can be used to indicate that the numbers of mixed marriages had been increasing for some considerable time prior to 1981. This can be inferred from the very much higher corresponding proportions among younger married persons in 1981, as illustrated in Appendix Table A4. This table shows that nearly 30 per cent of married persons aged between 13 and 34 years from the minority communities were in mixed marriages at that time. These proportions are broadly comparable with those estimated by Walsh in his 1970 study. The inferences regarding trends arc consistent with those observed by Bowen (1983) who estimated that mixed marriages increased progressively in the post war period, having been (in his view) of much less importance prior to that time.¹⁷

Mixed marriages are therefore very relevant to numerical loss in so far as the minority religious communities are concerned as the children of these marriages are disproportionately brought up as Catholics. These losses assume even greater significance if cumulative or intergenerational effects are taken into account. The evidence suggests therefore, that mixed marriages were a significant factor in contributing to the decline in the minority religious communities in the periods since the second world war. The impact of mixed marriages on the Catholic community (or, more precisely, on the number of Catholic children, which tends to be augmented) is minimal in relative terms, given that the number of homogamous Catholic families in 1991 was over 450,000 with an associated child population of nearly 1,190,000.

It is not clear, however, as to what alternative situation regarding the religion of children one might envisage in mixed marriage circumstances. One could consider, for example, a 50:50 subdivision of children in mixed marriages as reflecting a greater degree of equity in a religious context. In effect this practice existed to some degree in Ireland up to the early part of this century prior to the introduction of the Ne Temere decree. The custom sometimes prevailed whereby boys were brought up in the religion of the father and girls in that of the mother. If this criterion were applied to the 1991 data it would have the effect of raising the number of children affiliated to minority religions in mixed marriages from 4,000 to about 10,000, thus increasing the total number of such children from 37,000 to about 43,000.

Much of the emphasis in this section has been on the effect of mixed marriages on the recorded denominational affiliation of children and, in particular the numerical loss to the minority religious communities. It should be borne in mind that mixed marriages are now both an indicator of improved intergroup relations and a positive force for ecumenism, and can promote an enhancement of the religious life of many families. The issue of mixed marriages is deserving of more consideration than can be given here. Further research on this topic, using survey data, is currently being conducted by one of the authors of this report.

Return to list of contents

5
BIRTH AND DEATH RATES
FOR DIFFERENT RELIGIOUS GROUPS:
RATES OF NATURAL INCREASE

Having dealt with some of the principal factors which influence trends in the numbers in different religious communities, we are now in a somewhat better position to present and interpret a number of standard demographic indicators, i.e. birth and death rates and rates of natural increase.

BIRTH RATES

Turning first to birth rates (which consist simply of total births per thousand of the overall population in question), the relevant data are given in columns 2 to 4 of Table 8 which shows rates for Catholics and the minority communities for intercensal periods between 1926 and 1991. The absolute figures relating to actual births and aggregate population data on which these ratios are based are given in Appendix Table A.5. In a statistical sense the main technical difficulty here centred on categorizing total births according to religious groups for each intercensal period. Basically, the approach adopted was similar to that used in calculating the fertility indicators already discussed. For each census year in which questions on religious affiliation were asked the Catholic and minority communities proportions for the population aged 0 to 2 years was calculated and taken to represent the proportion which would relate to total births in these years. These proportions were then interpolated for the years between censuses and applied to total births in these years in order to give an annual series of births classified by religious group for the entire period from 1926 to 1991. The actual rates given in Table 8 are annual averages for each intercensal period, the relevant population denominators in each case being the average of the total population for each religious group at the beginning and end of the period.

The most noticeable feature is the low rates for the minority communities population when compared with the Catholics over the entire period in question. For the forty five years from 1926 to 1971, the Catholic birth rate exceeded that for minority communities by at least 50 per cent. The relative excess diminished slightly in the 1970s and there was a noticeable decrease in the 1981-1991 period when the relative difference declined to 36 per cent.

This outcome is hardly surprising in view of the results of the preceding analyses. The principal underlying factors derive from the relative fertility differences combined with the effect of mixed marriages. The higher rate of nuptiality among the minority population would have had an offsetting effect, but this was not large enough to have a significant impact.

One might have expected the birth rate for the minority communities to fill over the period in question in view of the decreases recorded in the fertility index as previously mentioned. The answer to this lies partly in the nature of crude birth rates. While the number of births in these communities did, of course, fall throughout this period (especially prior to 1961) the total population was also declining due to the combined effects of high numbers of deaths and outward migration. This served, in statistical terms, to maintain a stable birth rate, but at a very low level. For the Catholic population the sheer weight of the increases in the female population aged 15 to 44 years (especially after 1961) offset the decreases in fertility with the result that the birth rate for this community also remained constant (at a high level) for most of the period in question.

Apart altogether from their values relative to that for the Catholic population, the birth rates for the minority communities are extremely low by any standards, especially in the earlier periods under discussion. This feature was also noted by Walsh in his 1970 study International data for twenty one countries on birth rates covering the period from 1926 to 1951 contained in the 1954 Report of the Commission on Emigration¹⁸ do not reveal any countries with birth rates as low as those indicated for the minority communities in Ireland during this period. The lowest rate indicated in this source (14.7) was that for Belgium in the 1936 - 1945 period; most of the rates quoted were significantly higher than this. For later periods, the birth rates for minority communities are broadly similar to those prevailing in other countries. In the European Union (15 countries), for example, in the 1980s the overall birth rate was between 12 and 13 - a very similar figure to that indicated for the minority religious category in 1981 - 1991 given in Table 8.

Source:
See Appendix Table A.5.
Notes:
(1) For the period from 1926 to 1961 persons recorded as having 'no religion' and those who did not provide information on their religion are included with the 'other denominations' category. During this period the numbers involved were very small. For the period from 1961 these two categories are included with Catholics. See text for an explanation of these groupings.

DEATH RATES

The intrinsically low birth rates evident for the minority communities immediately raises questions as to whether these rates exceed the corresponding death rates. A relevant factor here is that the older age profile for the minority communities would suggest a relatively high death rate. This in turn, focuses attention on the natural increase (births minus deaths) for these communities. This is discussed in the next section.

Death rates for the two religious groupings in question are given in columns 5 to 7 of Table 8.¹⁹ The actual rates do, indeed, show that the mortality for the minority communities is significantly greater than that for Catholics, and that the relative differential has widened substantially over the years. In the 1926-1936 period the rate for the minority religious group, at 16.6 per 1,000 of the average population, was nearly 20 per cent higher than the corresponding rate for Catholics, which was 14 per cent. By the 1970s this differential had widened to over 50 per cent, even though it declined thereafter. The fact that such a differential existed as early as the 1920s would derive from the fact that even at this stage the minority communities were beginning to exhibit a growing distortion in their age profile, involving a preponderance of older persons. This was principally due to the heavy emigration which occurred within these communities between 1911 and 1926. Over succeeding decades the death rate for the entire Irish population (and by implication for the Catholic community) declined, mainly as a result of advances in medical and health care, the number reaching a level of 9.3 by 1991, compared with 14.2 between 1926 and 1936. These influences also applied, of course, to the minority communities, but the increasing proportion of older persons within this group exerted an offsetting influence and the death rate remained more or less constant until 1981. During the 1980s the death rate for this group fell noticeably to 12.5. In so far as one can determine, this was mainly due to a marked decline in the numbers in the 45 to 64 year age band between 1971 and 1981, a group that would have aged further in 1981-1991 and thus contributed to relatively fewer deaths. In other words the decline can be attributed to the onward movement in age of progressively smaller cohorts within the minority communities which had been depleted by emigration in earlier periods.

RATES OF NATURAL INCREASE

The death rates for the minority communities population exceeded the corresponding birth rates for all of the period covered except between 1981 and 1991. This gave rise to negative rates of natural increase i.e. apart altogether from any net gains or losses due to migration, these communities were not sustaining themselves in numerical terms (see columns 8 to 10 of Table 8). Throughout the entire period covered, the rate of natural increase for the Catholic population has been strongly positive, even though the total Catholic population declined on many occasions during this time span, or exhibited only moderate growth. because of heavy emigration. The negative aspect of the natural increase for the minority communities was particularly marked between 1926 and 1946. After 1971, however, the situation began to change and in the most recent decade between 1981 and 1991 the rate of natural increase for this religious group was positive, even if only very slightly so. This was mainly due to the fall in the minority communities death rate in this decade, as already discussed.

Return to list of contents

6
MIGRATION

SOME METHODOLOGICAL ASPECTS

Finally, let us consider the influence of external migration in so far as it has affected the numbers in different religious communities. Before any net migration trends are analysed it is necessary to comment briefly on the nature of migration estimates. The most reliable net migration estimates are those obtained for intercensal periods by linking changes in successive population stocks with the numbers of births and deaths occurring in these periods. This involves deriving aggregate net migration by subtracting the natural increase (births minus deaths) from the net change in population. This method cannot be directly used for deriving migration estimates for religious groups as the required vital statistics data are not disaggregated by religion. Another approach, the ‘cohort survivorship technique’, is to apply mortality or survivorship factors to the population of different ages at the beginning of an intercensal period in order to estimate the number of survivors at the end of the period.²⁰ The difference between the number of these ‘survivors' and the actual population figures in the corresponding age group in the next census can then be taken to represent net migration in relation to the age cohort in question. This approach can also be applied to births in the intercensal period in question and the aggregation of the net migration figures for all cohorts and categories yields a total figure. It will be noted that with both these approaches the final net migration total is estimated as a residual and thus bears the brunt of any errors and misclassifications associated with the process.

It is possible to derive net migration estimates by religion using the latter cohort survivorship technique as described. This can be done by applying the mortality factors to the age differentiated totals for different religious groups in the intercensal period base year and then comparing these with the corresponding age cohorts for different religious categories from the next census. One can apply the same procedure to births in the intercensal period as it is possible to compile reasonably accurate estimates for numbers of births according to religion (see section 5). This method does not, however, take account of differential mortality patterns between religious groups, as the same common factors are applied to each group. It is not considered, however, that this significantly distorts the final estimates.

An important feature to bear in mind in relation to these estimates is that their validity can be significantly affected by flows or changes of state other than migratory movements. The method can be deemed to yield satisfactory estimates of net migration as long as the religious groups in question remain reasonably self-contained, and there are not many transfers between these groups. However, if the scale of such transfers increases, then the situation becomes much more complex and inure difficult to interpret. In these circumstances the residual estimates reflect not only migratory movements, but also transfers between religious groups and it is not always possible to disentangle two influences. Broadly speaking, with Irish census data the issue of such transfers did not present a serious problem until the 1970s and, therefore, the above mentioned estimation method can be assumed to provide reasonable estimates of net migration for different religious groups between 1926 and 1971. The relevant data categorized according to religious denominations, are given in absolute terms in Appendix Table A5. Relative net migration data, expressed as annual average rates per 1,000 population, are given in Table 9.

After 1971 the situation became much more complex, mainly because of significantly increasing numbers in the category ‘other stated religions’ (which mainly involved a variety of smaller Christian groupings) as well as those who indicated that they had ‘no religion’ and those who refused to provide information on their religious affiliation in censuses. For these recent periods the formal methodology has still been applied, but interpretation of the results (set out at tile end of this section) is difficult and is, of necessity, partly judgemental.

MIGRATION TRENDS 1911 -1971

Let us first consider the period from 1911 to 1926 which, as already indicated in section 3, is especially important in view of the very significant events which it covered. Our earlier analyses indicated that the minority communities declined very rapidly during this time, from 327,000 to 221,000 down by 106,000 or 32 per cent. The corresponding relative decline for Catholics was very much lower at just over 2 per cent.

It is not possible to use the cohort survival method as described to estimate migration in the 1911 - 1926 period as simultaneous classifications of the total population according to religion and age are not available from censuses prior to 1926. This is unfortunate, especially in relation to a period when the minority communities population underwent a very substantial decline. This is not to suggest that certain features of the pattern of migration for different religious groups cannot reasonably be inferred from the available population trends, but it would have enhanced our analysis if more soundly based net migration estimates by religion for this period were possible.

It is of interest, nevertheless to try and estimate the decomposition of this large decrease into its constituent ‘natural increase’ and ‘net migration’ components. On the basis of the rates of the natural increase which applied to the minority communities in 1926-36 and in 1936-46 it can be broadly inferred that the natural increase associated with this group between 1911 and 1926 would have been about 10,000 (negative), bearing in mind that the period in question was of fifteen years duration. However, this figure would have been augmented (again in a negative sense) through deaths from the minority communities in the first world war. On the basis of data given in the 1926 census reports, these can be roughly estimated at about 5,000²¹ implying a total ‘adjusted’ natural increase of about - 15,000. This would suggest therefore that actual net emigration for the minority communities from what is now the Republic of Ireland between 1911 and 1926 was at least 90,000.²² A significant proportion of this outflow would, of course, have been due to the withdrawal of British security forces and their dependants in early 1922. Relying again on estimates based on figures quoted in the 1926 census reports one can estimate that the size of this outflow (i.e. that part of it which related to the minority religious groups) was about 30,000.²³ This would thus imply a net outflow of about 60,000 from the ‘indigenous' minority communities between 1911 and 1926. This is equivalent to a rate of between 14 and 15 per 1,000 of the average population, a much higher figure than that which applied to Catholics during this period, which was between 7 and 8. This rate is also higher than those for the minority communities in the following decades, which were between 9 and 10 (see below).

Source:
See Appendix Table A5.
Note:
For the period from 1926 to 1961 persons recorded as having 'no religion' and those who did not provide this information are included with the 'other denominations' category. During this period the numbers involved were very small.
For the period from 1961 to 1971 these two categories are included with Catholics. See section 2 for an explanation of these groupings.

As this estimate may even understate the net outflow in question, there would appear to be little doubt therefore that this transitional period witnessed a very considerable exodus of persons from the minority religious communities in Ireland. One must reasonably assume that the main causation underlying this exodus was a sense of apprehension in certain sections of these communities associated with the transition to national independence in 1922 and the upheavals which attended that event. Others may have concluded that their future prospects were not as advantageous as they were previously. This proposition appears to be borne out by the continuing high emigration for this group in the period after independence (as indicated in Table 9) and the fact that,while most of the emigrants involved were young, they also included significant numbers in the older age groups. Furthermore, the parallel occurrence of relatively substantial emigration in age groups covering those aged less than 15 years, and of persons aged between 25 and 44 years (see Table A6) suggests that the outflow involved many family groups. Support for these views can be found in a number of the other studies of the religious minorities in Ireland, especially those of Bowen (1983) and Kennedy (1973).

In the period up to the end of the second world war a still significant, if reduced, degree of emigration continued to apply to the minority religious communities. The relevant migration rates reached as high as 10 per 1,000 population, again markedly higher than the Catholic rates which were between 5 and 6 over this period (see Table 9). Bowen (1983) expressed the view that the more advantageous social profile of the religious minority communities, and their significant involvement in the commercial and business life of the country, should have led to less rather than relatively more emigration vis á vis Catholics during this time. While these continuing high outflows may be partly explained by some residual effects attributable to the political transition of the early 1920s, new legal measures introduced in the immediate post - independence period are also likely to have played a role. Throughout the 1920s, the new Irish Government introduced a series of laws which, in the view of the minority communities, would have reflected a strong Catholic of nationalist ethos. Most notable were measures prohibiting divorce, banning the sale of contraceptives, censorship of publications and films and the introduction of compulsory Irish tests in school examinations and in entry to public sector employment. While some of these measures must be considered as obvious and predictable consequences of the political transition (e.g. the promotion of the Irish language), the whole range of changes is likely to have been viewed at the time by the minority communities as moving towards a form of society in which they would not be completely at ease. The evidence suggests that in these circumstances many in the minority community opted to emigrate. A more detailed discussion of such issues in so far as they relate to this period is contained in Kennedy (1973, p. 12).

The situation changed in the post World War II period. The incidence of emigration for the minority communities decreased somewhat and Catholic emigration rose greatly, to a much higher level (12.4 per 1,000 population) than that which existed for the minority community. In this period. especially in the 1950s which involved periods of deep recession, the evidence suggests that economic considerations, which applied to all sections of’ the population, were the main influences which gave rise to very large migratory outflows. In the 1960s, with the advent of increased economic growth, overall emigration diminished. To the degree that it did exist, our analysis suggests that it applied equally to the Catholic and minority religious communities. Table 9 shows that between 1961 and 1971 virtually identical rates of net emigration prevailed for both communities, at about 4.5 per 1,000 of the average population.

ESTIMATING MORE RECENT TRENDS 1971-1991

The estimation of migration trends for different religious groups for recent decades is much more difficult. The emergence during the 1970s and 1980s of a growing tendency to change religious affiliation, or to withdraw from membership or adherence to any form of religious worship (i.e. no religion) means that the residual ‘net migration’ estimates derived from the above mentioned methods in effect represent a combination of migratory movements and transfers between religious groups (including, of course the ‘no religion’ category). The position is further complicated by the increasing propensity for persons not to reply to the census questions on religion.

Notwithstanding these difficulties, it is still of interest to apply the net migration estimation procedure to the data for recent decades and to attempt to interpret the outcomes. There is, to some degree, an identifiable pattern to the nature of the flows and transfers involved. The results are shown in Table 10 for the 1971-81 and 1981-91 intercensal periods, in these analyses, in order to provide better insights, the religious groupings have been extended somewhat compared with earlier tables. The three main Protestant denominations (Church of Ireland, Presbyterians and Methodists) have been distinguished as a separate group, with the remaining religious groups (other than Catholics) classified under the ‘other stated religions’ category. Those who were recorded as having ‘no religion’ are also separately distinguished, but persons who did not supply information on their religion in censuses are included with Catholics.²⁴ The residual ‘net migration/transfers’ estimates relate only to the population in the base years, 1971 and 1981. i.e. no account is therefore taken of flows or transfers relating to those who were born within each of these periods. The relevant data are given in absolute terms and as rates per 1,000 population for each intercensal period.

Looking first at the estimates for 1971-81 (which, it will be recalled, was a period which involved a substantial overall net migratory inflow these show that the relative indicator for the Catholic community, at 1.9, is significantly less than the national rate of 2.9, while that for the main Protestant denominations is actually negative (-1.4). These figures suggest a loss for each community to the other religious groupings (including no religion). While the relative losses for the main Protestant denominations would appear to be greater view of the negative rate in an overall period of net inward migration), the actual losses for Catholics in numeric or absolute terms would have been larger in view of the predominant size of this group. The estimation procedures indicate very large relative transfers to the ‘other stated religions’ and ‘no religion groups', especially the latter. If one excludes the possibility of actual migration, the estimates suggest that some 30,000 persons recorded in the ‘no religion’ category in 1981 were included as members of specific religious groups in 1971.

Between 1981 and 1991, a period characterized by heavy net outward migration, the rate of net loss for the Catholic group, -7.0, is greater (in a negative sense) than the national rate of -5.5, while that for the main Protestant denominations is somewhat lower at -3.6. One can tentatively interpret these figures as suggesting that the Catholic community lost members both through emigration and transfers to other groups, while the losses for the main Protestant denominations may have been mainly through outward migration (if one assumes that the pattern of migration applying to this group was broadly the same as that which existed for the population as a whole). As in the 1970s, it is clear from the very large rates of emigration/transfer that the 'other stated religions’ and ‘no religion’ categories were again augmented by significant movements from other religious groups. The evidence suggests that in this decade these inflows must have consisted predominantly of Catholics. A notable difference in this period (vis á vis the 1970s) was the large influx into the ‘other stated religions' category.

Notes:
(1) While the 'migration/transfers' estimates for different religious groups relate both to migratory movements and transfers between religions, the estimates for the total population relate only to net migration.
(2) The estimates in this table relate only to the base year population (i.e. for 1971 and 1981) and take no account of net flows relating to those born in each period.
(3) The rates per 1,000 population are calculated on the basis of the base year populations for 1971 and 1981.
(4) The main Protestant denominations category includes the Church of Ireland, Methodists and Presbyterians.

Return to list of contents

[Continued...]

9. In 1971 extramarital births comprised only 2.7 per cent of all births, but by 1991 this percentage had risen to 16.6.
10. Calculations relating to marital fertility are not possible for the censuses prior to 1946 as tabulations involving classifications by religion and marital status were not given in the census results.
11. See Appendix I for further information on this issue.
12. Simulations with these age structures (using constant age specific fertility rates for all women for 1991) would suggest that the slight differences winch exist in the age distributions between the two religious groups would have the effect of raising the Catholic fertility level by some 4 per cent when compared with that for other denominations.
13. Similar calculations are not possible for earlier years as tabulations involving marital status and religion were not compiled for censuses prior to 1946.
14. Prior to the Papal decree Ne Temere in 1908 the Catholic Church recognized mixed marriages as valid even if they were not celebrated before a Catholic priest. While it is believed that some Catholic priests demanded a promise of the partners that any children would be baptized and educated as Catholics before they would solemnize a marriage, it was not always insisted upon.
With Ne Temere mixed marriages were treated like marriages involving two Catholics in that to be valid in the eyes of the Catholic Church they had to be contracted before a Catholic priest. Before a Catholic priest would solemnize a marriage between a Catholic and a person of another faith a dispensation had to be obtained from the Church from the impediment of it being a mixed marriage. Catholic Bishops would not grant a dispensation from the impediment until both partners promised to baptize and educate the children as Catholics. In this way the promises became necessary for validity.
The Vatican’s apostolic letter of 1970, Motu Proprio, represented a significant relaxation of the conditions of Ne Temere. It declared that as a condition of a dispensation being granted from the impediment of a mixed marriage only the Catholic partner need promise that she/he do all in her/his power to have all the children baptized and brought up in the Catholic Church.
15. See section 2 of this report regarding the coverage of household members in censuses.
16. Since the minority communities (especially in recent years) included significant numbers in the ‘other stated religions’ category (see Table 1(A)) there is a possibility that the number of Catholic children in mixed marriages could be overstated. This could arise, for example, if one spouse in a family was recorded in the census as a ‘lapsed Catholic’ and the other as ‘Catholic', as such a marriage would have been classified as ‘mixed’. It is likely that all the children in such a family would have been returned as Catholics. The position of those who recorded themselves as ‘Christian’ might also be unclear in these circumstances. The numbers involved. however, are relatively small and unlikely to affect the pattern of the results as given.
This can be reaffirmed by considering only mixed marriages involving Catholics and Church of Ireland members (the largest of the minority religious groups). In such marriages the proportion of Catholic children in 1991 was 73 per cent, almost as high as the proportion for all mixed marriages given in Table 6. Compton and Coward (1989) using a small sample (n = 148) of mixed marriages in Northern Ireland also found that the majority of children within these marriages were recorded as Catholic (56 per cent, as against 37 per cent for other denominations).
17. There is some evidence which raises a query regarding this proposition. Maguire (1993) in analysing the situation of working class Protestants in three districts in Dublin city in the early 1900s found that some 15 per cent of Protestant males in this class were in mixed marriages with Catholics. It is possible, of course, that the incidence of mixed marriages may have declined after the early 1920s, given that this was a period of greater intercommunal tension. An interesting figure of Maguire’s results is that the proportion of Catholic children in these earlier mixed marriages was remarkably similar to the figures revealed by the present study, just under 80 per cent.
18. Commission on Emigration and other Population Problems, Reports, 1954, Table 67, p. 90.
19. The derivation of these rates, which involved a great deal of background computation, is explained in Appendix II.
20. This is basically the same method as that adopted in estimating the number of deaths by religious group as explained in Appendix II.
21. The reference in this case is the general report on the 1926 census of population (Volume X), page 12. Quoting Registrar Generals’ data this source indicates that the number of Irish warrant officers, non-commissioned officers and men who died outside of Great Britain and Ireland from 1914 to 1918 was 27,400. The number of Irish officers killed is not recorded. If one assumes that about one half of the above total were from the counties of the Republic and that, in turn, one third of these were from the minority religious communities, this suggests a war deaths figure of about 5,000.
22. This is probably a minimum figure as the estimated natural increase in this period may riot have been as large (in a negative sense) as - 15,000. It is likely that the net migration associated with the minority communities prior to 1911 would not have been uncharacteristically large in relative sense and would not, therefore, have contributed to serious imbalances in population structure (with regard to age, for example) which would, in turn, have led to undue reductions in the natural increase.
23. See the general report of the 1926 census (Volume X), p. 47. This source indicates that, according to the 1911 census, the number of British army, navy and police personnel in Ireland in 1911 who were members of minority religious troops was 21,400. If one assumes that there were 37 dependants for every 100 such personnel (the known proportion for the British army in Dublin) this would suggest a larger group of about 30,000, the great majority of whom would have left the country in 1922.
24. The reasons for this are set out in section 2.

Return to list of contents

APPENDIX TABLES

Source:
Special tabulations from the 1981 census.
Notes:
(1) The 'mixed marriage rate' for males is the number of males from the minority communities in the appropriate age group who were in mixed marriages taken as a percentage of all married males from these communities in the same age groups. The rates for females and persons are similarly defined.
(2) As in the other tables in this report, the minority communities category includes the main Protestant denominations and 'other stated religions', but excludes 'no religion' and those who refused to provide information on religious affiliation in the census.