Equality and Equity Report

CCRU home background on CCRU community relations equality and equity research

Studies in Employment Equality
Research Report No 3


Ian Shuttleworth

School of Geosciences
Queen's University, Belfast

(December 1994)


4.1 Introduction

Religious differentials in labour market outcomes in Northern Ireland are well known (see for example Whyte 1990; Cormack and Osborne 1991; Bardon 1992; Murphy and Armstrong 1993) Most quantitative research (Compton 1991; Smith and Chambers 1991) has concentrated on the adult labour market, and particularly on the incidence of differential experience of male unemployment which has persisted for a considerable period.

Less well documented are the means by which these inequalities are produced and the relationships between religion, schooling and early careers, Despite this lack of quantitative evidence, it is highly likely that the experiences of young people in this period of transition from school (to either economic activity or other forms of education) are very important. Recent research in other areas (Breen 1991; Narendrenathan and Elias 1993) has suggested that this transition stage has a strong influence on later life histories. Early career experiences for Northern Ireland young people can therefore be reasonably expected to be of major importance in creating religious differences in later life.

The main question is not so much who gets work or becomes unemployed in this age group. This picture of youth transitions from school to either work or unemployment, is now outdated by two factors, Firstly, the rate of participation in post-compulsory education has increased in all parts of the UK (none more so than in Northern Ireland) . Recent evidence estimates that 45% of young people remain at school past the age of 16 (Training and Employment Agency 1993) ; this 45% does not include young people who attend colleges of further education and thereby increase age-participation rates in post-compulsory education,

Secondly, the government, via programmes such as YTP has intervened evermore vigorously in youth transitions so an increasing proportion of school-leavers have participated in these schemes.

This means that young people are not confronted with a simple choice on leaving school of either work or unemployment. There are now a number of options they might enter, For a minority, the traditional path from school to work still exists. For the majority, however, transitions from school are marked by spells on training schemes and for yet others they are delayed until the ages of 17 and 18 and the completion of post-compulsory education.

The reasons for these changes are unclear. Negative factors such as the disproportionate rise in youth unemployment in the UK since 1970 have been identified (Hart 1988; Ashton et al 1990) which has meant that youth training has been used as a palliative measure (Finn 1987). On the positive side, however, the expansion of educational provision has resulted partly from a lack of other choices for young people but also from government policies to increase educational provision.

4.2 The status of young people aged 16-24

Because of these considerations, it would be useful to show what is known about the extent of inequality by religion among young people in Northern Ireland in terms of their participation in education, training, employment and unemployment both as something of interest in itself and as a context for the analysis of the SELS. Whether or not there is agreement on the causes of these changes, it is plain that they have had major implications in Northern Ireland. Analytically, they mean that a narrow approach that seeks to analyse employment or unemployment alone fails to come to terms with the impacts of changes in training and post-compulsory education. In looking at religious differentials in early career experience they also mean that there are alternative ways in which inequality can develop, rather than just by means of differential access to employment.

The data source used for this analysis is the Northern Ireland Labour Force Survey (NI LFS). This was an annual government survey of some 4,000 Northern Ireland households carried out in the spring of each year. The data used in Charts 4.1 to 4.5 and Tables 4.1 and 4.2 are taken from a pooled sample of data collected in 1985-86 and 1990-91. The data therefore do not refer to any single year. Furthermore, the data are not longitudinal in that it is impossible to trace, for example, the fortunes of the young people aged 16 in 1986 at age 20 in 1990, Instead the data are cross-sectional in that they provide a snapshots of the status of the young people at the times of the surveys, For this section of the analysis 16-24 year olds are the subject. This broad age group was chosen because it covers the full range of contemporary elongated youth transitions into which the SELS school-leavers are introduced on leaving school. The SELS, of course, does not cover the full range of the NI LFS data, Instead it refers only to the first year after leaving school and is therefore a sub-sample of young people aged over 16.

In interpreting the data, two caveats should be made. Firstly, the NI LFS may under-represent certain types of young people. As a survey of households in Northern Ireland, it may underestimate the proportion of young people aged over 18 in education. This is because it does not account for students in higher education who have taken places at institutions in Britain. Given the probable dominance of Protestants in this migration flow (Cormack and Osborne 1993). this may mean that Protestant participation in higher education for 18-24 year olds is not properly estimated.

Secondly, the NI LFS is a sample survey. Because of this, estimates of population characteristics are subject to sampling errors. This is especially the case when small numbers are involved (for examples economically inactive males aged over 20). In these cases inter-religious differences are within the range of sampling error. Despite the need for caution, the NI LFS provides a useful summary of trends in youth transitions in Northern Ireland; the trends are plain. Confidence in the general patterns being observed is strengthened by similar trends noticeable in other sources such as the SELS, the Continuous Household Survey (CHS) and the 1991 Census of Population.

Charts 4.1 to 4.5 should be viewed together to gain an overall picture of the status of young people. In keeping with the remarks made about the necessity of considering youth transitions as a whole, the participation rates for each state are calculated as a percentage of the total Population in each year age group. Chart 4.1 shows employment rates; the feature that stands out is the higher Protestant employment rates, Charts 4.2 and 4.3 suggest some reasons why this might be the case. Looking first at participation in government training schemes, Catholics consistently have a higher proportion in training than Protestants, particularly at the age of 17. The unemployment rates revealed in Chart 4.3 are intriguing For 16 and 17 year olds, the proportion of unemployed Catholics is about the same as that of Protestants. For those older than 17, Catholic rates of unemployment appear to increase faster than Protestant rates until they are about twice as high. Chart 4.4 shows participation in education. Generally,, the participation of Catholics in education is higher than f or Protestants, though for 18-24 year olds there is some doubt about the accuracy of this assumptions. Finally, Chart 4.5 suggests that Catholics are much more likely than Protestants to be economically inactive (excluding those in education).

These overall presentations of the situations of Young Catholics and Protestants conceal interesting differences once gender is taken into account. Tables 4.1 and 4.2 accordingly present similar data for males and females respectively. Again, the two tables should be considered together. just looking at the experiences of each gender, they show that males of both communities have higher fates of participation in employment and government training. At the same time, females of both religions have higher rates of participation than males in education and other economic inactivity, There are thus many features that are in common between Catholics and Protestants in the general shapes of youth transitions

Table 4.1: The Status of males aged 16-24 by religion (percentages)


Source: NI LFS n=3194

Table 4.2: The status of females aged 16-24 by religion (percentages)


Source: NI LFS n=3290

However, differences creep in once participation rates are considered by gender and religion. For males, the contrasts between Catholics and Protestants are especially marked. As a general feature, Catholic males aged 16-18 have higher rates of participation in government training and post-compulsory education than Protestants. To balance this, Catholic males have lower rates of employment. For males aged over 18, the religious differential in unemployment rapidly increases so that Catholic males have eventually an unemployment rate about two times higher than their Protestant counterparts. For females, the picture is not so clear. Catholics younger than 18 tend to have lower participation rates in post-compulsory education than Protestants. At the same time, those older than 20 generally have higher rates of economic inactivity (excluding education) than their Protestant counterparts.

The first observation to make about these patterns is that the processes producing religious differentials in the labour market were still operating in the late 1980s and early 1990s. Differentials amongst older age groups might be dismissed as being a product of inequities in the Northern Ireland education system and labour market in the 1940s and 1950s and thus, in some sense, an historical relict (Eversley 1989). Well-documented historical inequities (Bardon 1992; Smith and Chambers 1991) cannot explain contemporary processes which are differentiating new entrants to the labour market and are influencing contemporary school leavers.

It is difficult to explain these patterns. They need not necessarily reflect direct discrimination on the basis of religion amongst young people. Structural explanations of religious differentials in unemployment rates and occupational structure have been well-rehearsed (Compton 1991; Smith and Chambers 1991) and could just as easily apply to young people as well as adults.

However, this type of explanation refers mainly to adult unemployment. For young people, something more appears to be happening, particularly for males. The main feature, especially for those aged 16-18, is a lower Catholic employment rate and a greater participation in education and training. once this 'scheme effects ceases at the age of 18, unemployment differentials begin to climb steeply. Two competing hypotheses, with very different policy conclusions,' can be used to rationalise this greater participation in education and training. The first is that Catholics,, because of a history of high relative communal unemployment and because they live in areas of high unemployment, are 'discouraged workers' (Raffe and Willms 1989). In its simplest form this approach suggests that young people are delaying entry to the labour market by continued participation in education because they expect to have a higher chance of becoming unemployed. A modification of the hypothesis could be used to explain higher training participation. The second argument is that Catholics are seeking to accumulate human capital so they may compete in what they rationally expect to be a tough labour market.

At present, there is no information that allows a rational choice between these two hypotheses. However, there is strong evidence that recent changes in the structure of the youth labour market and youth transitions, have not altered the balance of religious inequality in Northern Ireland. The expansion of education and training does not appear to have redressed Catholic-Protestant differentials,, but instead has merely replicated them in different forms.

This is the context to the analysis of the immediate post-school experiences of recent Northern Ireland school leavers. The main direction of the analysis of the SELS will be to investigate the extent to which religious inequalities amongst young people occur and to consider how far they might be explained by structural factors such as social background and educational attainment.

4.3 The SELS data on first destinations and analytical approaches

The SELS, besides data on qualifications, has information on the status of school-leavers between 9 and 12 months after leaving school. Five categories were defined to describe the status of school-leavers; employment, government training, unemployment/other economic inactivity, further education and higher education.

Table 4.3 shows the distribution of school-leavers between these states, by form from which the school-leavers left, gender and religion. The same type of patterns as seen in the NI LFS are immediately noticeable. These are lower rates of participation in employment rates for Catholics, but greater Catholic participation in training and education. For all males, the differences are statistically significant at the 0.05 level; the differences between males are also statistically significant at the same level for fifth form and lower-sixth form leavers. For females and for upper-sixth form male leavers differences are statistically insignificant. This suggests that the major source of inequalities by religion lies among males, and especially males who leave from fifth and lower-sixth forms.

Table 4.3: Religious inequalities in the BELS by form from which left and gender (percentages)

All leavers[1]
Further education
Higher education
Fifth form and lower sixth leavers[2]
Further education
Higher education-- --
Upper sixth form leavers[3]
Further education
Higher education

Source: SELS

Note 1: For males, n=745; for females, n=760
Note 2: For males, n=458; for females, n=452
Note 3: For males, n=283; for females, n=303

The patterns of inequality we see do not exactly follow those seen in the NI LFS. This is because the SELS was drawn from a different population; instead of young people in households, young people who had recently left school. This explains in part the differences between the two datasets. Also, because the SELS, is confined to school-leavers, it omits an important part of the experience of young people, that is to say those who stay on at school. Despite this, the SELS has many interesting features and the patterns of religious inequality observed accord with those seen in other Northern Ireland sources, As a sample of young people who have recently left school, the SELS database is useful as it can be used to show how inequalities between Protestants and Catholics develop and the extent to which the religious differentials arise within a year of leaving school.

The type of analysis that will be pursued seeks to discover how differences in immediate post-school destination can be explained by structural differences between Catholics and Protestants as measured by family size, educational qualifications and household circumstances. The form of the analysis is thus very close to that seen in Chapter 3. In the Northern Ireland context, the approach is therefore similar to those of Smith (1987), Compton (1991), Smith and Chambers (1991) and Murphy and Armstrong (1993) in that variables which measure SES and family circumstances are taken into account within models to see how far religious differentials persist. The argument, in summary, is that religious differences amongst young people may be accounted for by different rates of qualification and varying household circumstances between the two communities rather than by religion alone and are thus an artifact of the data that is re-moved by standardisation, Though the basis of this approach appears to be generally agreed, in practice there are considerable difficulties in interpreting and understanding the results of these analyses as is witnessed by often lengthy disputes (Cormack and Osborne 1989; Compton 1991; Smith and Chambers 1991).

Outside Northern Ireland this type of survey-based analysis of the behaviour of young people has a long pedigree. Main and Raffe (1983) investigate the incidence of unemployment and employment among young people in Scotland among 1979 school- leavers. Their model had a wide range of explanatory variables; family size, occupational class of father, education of parents, whether or not respondents had sat O grades, whether or not the respondent had severe truancy problems in the last year at school, the denomination of school (whether or not Roman Catholic) , whether or not a part-time job had been held at school and local unemployment rates. Main and Raffe (1983) found, of these, that three variables appeared to be important, Firstly, having sat O grade exams increased the chances of being in employment for both males and females. Secondly, those who had had a part-time job at school had a significantly greater chance than those who had not of getting a job after leaving school. Finally, local unemployment rates were significant with higher local rates depressing the probability that individuals would be in employment.

Payne (1987) , using 1980 and 1981 data from the General Household Survey (GHS) investigates the effects of patterns of household economic activity on the fortunes of economically active 16-19 year olds. The same type of explanatory variables were used as selected by Main and Raffe (1983), since, as Payne (1987) comments, these might be assumed or have been shown to be associated with youth unemployment. These were; regional unemployment, qualifications, age, age of leaving full-time education, sex, ethnicity, illness, occupational class of head of household and economic status of other household members. The findings of this research are very clear; unemployed young people are more likely than their working peers to have another family member out of work or economically inactive. Qualifications gained and regional unemployment rates are also significant determinants of unemployment.

Several lessons can be learnt from these British studies. Firstly,, and most obviously of interest,' is the failure of religious denomination of school to be significant in the Scottish study of Main and Raffe (1983) once other social characteristics had been controlled for in a model. Secondly is the general agreement that qualifications, local labour market conditions and household circumstances are important determinants of the probability of being employed or unemployed The final point concerns the dates of the datasets. Both these datasets date from the beginnings of the 1980s and so pre-date the changes in youth transitions caused by government education and training policy (most conspicuously the expansion of training places and the alteration of the rules governing access to state benefits for young people in 1988). This means that the simple dichotomy between employment/unemployment as discussed by Main and Raffe (1983) and Payne (1987) is now somewhat outdated.

4.4 Method

To handle the variety of experiences in contemporary youth transitions a more complex analytical structure was used than in previous analyses (see Appendix 3 for further detail) . In summary,' the analysis is conducted for two separate sub-groups of the SELS because of the empirical patterns of post-school destinations observed in the data; fifth-form leavers, for example, have a negligible chance of immediately entering higher education. Because of this age-dependent structure, the main distinction was made between upper sixth-form school-leavers, and those who left from fifth-form and lower sixth-form as this seemed the most logical means to analyse the data.

To cope with the structure of the SELS data on first destinations in terms of the patterns of education, training employment and unemployment, the analysis had to take account of a much more complex pattern than a simple dichotomy between employment and unemployment. Accordingly, additional analyses were undertaken. For both analytical sub-groups, the first stage was to analyse the determinants of economic activity (defined as employment, training and unemployment) and economic inactivity (further education, higher education and other economic inactivity) . The second stage then involved a set of subsidiary analyses within each sub-group. Thus, to take an example for fifth-form/lower sixth school-leavers), the determinants of employment would be analysed for those who were economically active. The structure of the analysis is formally presented in Appendix 3; but the structure will become apparent from the direction of the argument.

As the outcome variables were coded as discrete categories, logistic regression was used as the most suitable and appropriate I technique. The explanatory variables were much the same as those used in Chapter 3 but with some minor differences in coding. The main additions to the list of explanatory variables were a series of measures of GCSE and A Level passes by subject type (passes in Science, Mathematics and English). The implications of the use of these variables will be discussed later in more depth.

4.5 Staying-on and leaving school

The analysis begins with an attempt to estimate some of the determinants of staying-on at school. In the absence of better data and more fully-specified models, it provides some ideas for further research and outlines patterns of data within the SELS. Since the SELS does not have data on young people who remain at school past the age of 16 , the best that can be done is to estimate the effects of personal/family characteristics on the probability of being a fifth-form leaver (rather than leaving from sixth form) . This approach assumes that relationships between social processes and education did not change markedly in the two years separating the fifth-form leavers of the SELS and the time that upper sixth-form leavers were in the fifth form and able to leave school.

Models [10] were estimated separately for grammar-school and secondary-school pupils. These are shown in Table 2 of Appendix 3. There are some contrasts in the results between the two types of school. In secondary schools, males were less likely to stay on than females as were those eligible for free school meals (a proxy for Income Support). On the other hand, Catholics were more likely to stay-on than Protestants. In the grammar-school sector, males were again less likely to stay-on than females. Those who were eligible for free school meals were also less likely to stay-on though not significantly so. The religion effect, however, was insignificant.

What can we make of these findings? The first observation is that the patterns in the SELS data are very much like those seen in the NI LFS. These are that females are more likely to remain in education than males, and the suggestion that Catholics are more likely to remain in education than Protestants. The second observation is that there seems to be a differential relationship between religion, schooling and the decision to stay at school between the grammar-school and secondary-school sectors. This may mean that markedly different educational processes, relationships with the labour market and attitudes to education exist by religion and type of school last attended, it is difficult to gauge the implications of these potential differences but they form an interesting area for further research since this is as far as the analysis can be taken with the SELS.

4.6 Fifth-form and lower-sixth form leavers

4.6.1 Economic activity versus economic inactivity

The first model estimated for this group of school-leavers examined the influences on the probability of an individual being economically active rather than inactive (entering a college of further education or withdrawing from the labour market and the education system). The statistically significant variables were firstly age; older respondents had a greater chance of being economically active. Secondly, gender, with males having a greater likelihood than females of being economically active. Thirdly, religion was statistically important as Catholics were less likely than Protestants to be economically active. Fourthly, three qualification variables were statistically significant. in general, school-leavers with high numbers of GCSEs were less likely to becoiae economically active as were males with GCSE passes (at grade C or higher in Mathematics) . Finally, males with relatively many siblings were more likely to be economically active than other respondents (see Table 3 in Appendix 3).

4.6.2 The economically active: employment

Within the sub-set of economically active, it is clearly important to understand the determinants of entry to employment, training and unemployment. The results of these analyses are considerably more complex than those reported earlier in the chapter. Because of this only summaries of the main findings will be presented in the text (though as before the model coefficients will be tabulated in Appendix 3).

Looking at employment, there are three results that are easily interpreted. These are that the chances of employment for the economically active increase with age, with the number of qualifications and if an individual is male. This simple picture is immediately complicated because of the presence of a number of complex interaction terms with religion.

These mean that employment chances tend to increase if the respondent is a male (verging on statistical significance) unless they are a Catholic male (when they tend to decrease) Likewise, if an economically active fifth/lower-sixth form school-leaver has both parents employed, their chances of employment decrease if they are a Catholic. A large family effect (associated with religion) is significant as Catholics from large families have a lesser chance of being employed than other respondents (see Table 4 in Appendix 3 for the full models for YTP, employment and unemployment).

4.6.3 The economically active: training

The pattern is as complex once training participation is considered. Here increasing age has a negative effect on the chances of being in training as would be expected given the peak in training participation for 16 and 17 year olds. Also, given what is known about the qualification level of young people on YTP (Shuttleworth and Armstrong 1993), it is unsurprising to note that a growth in the number of GCSEs held by an individual is related to a decrease in the probability of being in training. However, there are also a number of interesting interaction effects involving religion. These can be summarised as follows. Firstly, males have less likelihood of being in training unless they are Catholic males. Secondly, there is a social deprivation and religion effect, as measured by eligibility for free school meals, that means Catholics from deprived backgrounds are more likely to be in training than other leavers. Finally, as a mirror image to the effects noted for employment. Catholics with many brothers and sisters, and Catholics with both parents, or only their father in work, or only their mother in work, are more likely to be in training than other school-leavers.

4.6.4 The economically active: unemployment

The estimates for the relationship between personal/family characteristics are much easier to summarise and interpret. Basically, the main item of interest is that school-leavers with both parents in employment are significantly less likely to be unemployed than other young people. This seems reasonable in the light of other research findings (Payne 1987).

4.6.5 The economically inactive: further education

Most (about 94%) of the economically inactive enter colleges of further education. For school-leavers of this age, economic inactivity is very nearly synonymous with entry to a further education college. However, to complete the summary of this section of the analysis, it was thought worthwhile to estimate the effects of social characteristics on the probability of an economically inactive school-leaver being in further education. The main effects increasing the probability of being in further I education are having both parents in work, being Catholic and having GCSES. School-leavers from larger families tend to have a smaller chance of entering further education.

The religion coefficient is very interesting coming, as it does,' as part of a series of coefficients in different models suggesting, everything else being equal, that Catholics have a greater likelihood than Protestants of participating in post- compulsory education and training (see Table 5. Appendix 3 for the model coefficients).

4.7 Upper sixth-form leavers

4.7.1 Economic activity versus inactivity

The same procedure f or these leavers was followed as for the fifth-form/lower sixth group. The first stage of the process therefore estimated the influence of personal/family characteristics on the probability of being economically active. The results of the model were simple to interpret as there was only a single interaction term though it again involved religion. The largest effects, reducing the chance of being economically active, was the possession of qualifications at both A Level and GCSE. What this meant was that school-leavers with higher numbers of these qualifications were more unlikely to enter the labour market than those lesser qualified. Catholics, unless they were male, tended to have a smaller chance of becoming economically active; if male they tended to have a greater probability of entering the labour market (see Table 3 in Appendix 3).

4.7.2 The economically active: employment

Of those who entered the labour market, three factors stood out as being statistically significant in influencing the probability of being in employment. Firstly, qualifications; those school- leavers who had A Levels had a greater chance of being in employment. Secondly, parental labour market status with school- leavers with both parents (or only their father employed) having a greater chance of employment than other respondents in this sub-group. (The direction of this relationship was as expected given the result of the British studies cited earlier.) The religion coefficient was statistically insignificant (see Table 6 in Appendix 3 for more details of the model).

4.7.3 The economically inactive: further education

Further and higher education are the main destinations for upper sixth-form school-leavers as most become economically inactive. Of these, higher education is dominant. In estimating the effects of personal/family characteristics on being in further education, several trends emerge that are consistent with those seen elsewhere in the SELS. Some of the effects are intuitive and obvious. The chances of being in further education diminish as the number of A Levels and, to a lesser extent GCSES, a school-leaver has increases. This simply means that upper sixth- form leavers with these qualifications enter higher education. A number of complex effects involving religion and gender emerge. Overall, males are less likely to enter further education (as are Catholics) but a strong male/religion effect is also apparent. Two counterbalancing social deprivation effects appear in the results; if a respondent had been eligible for free school meals, they were less likely to enter further education than someone who had been ineligible. On the other hand leavers who had more brothers and sisters were more likely to be in further education than those from smaller families (see Table 7 in Appendix 3 for more details on the models on the incidence of further and higher education).

4.7.4 The economically inactive: higher education

The determinants of being in higher education do not include religion. As one might expect, school-leavers with A Levels have a greater chance of being in higher education than those with no A Levels. The A Level squared term is interesting; what it means is that the relationship between the number of A Level gained and the chance of being in higher education is non-linear. So a leaver with three A Levels has not got three times the chance of being in higher education than a school-leaver with one A Level but perhaps, for example, nine times the probability. Some interactions with gender are noticeable but it is unclear exactly how they might be interpreted. Males have a lesser chance of being in higher education than females unless they have many GCSEs and are older. Finally, those with an A Level pass in a science subject seem to have a greater chance of being in higher education than school-leavers without an A Level Science pass.

4.8 Discussion

Looking at the immediate post-school destinations of young people, there are many interesting effects that are worthy of further comment. These include the influence of family background, in terms of the labour market status of parents and experience of social deprivation, and gender. To reduce the discussion to a manageable size, these will be largely ignored though in discussion of the importance of religious background that follows their importance is implicit.

4.8.1 General comments about religious background

The SELS data show a number of interesting effects associated with religion. A number of the other effects noted in the data may be 'nonsense' correlations reflecting sampling variation and the structure of the data. However, the influence of religion as noted in the models is sufficiently strong and consistent to allow considerable confidence in its existence.

For this sample, Catholics have, everything else being equal, a greater chance of remaining in education (economically inactive) than Protestants, once economically active they have, other factors being equal, a greater chance of being in training. These effects are particularly strong for those who left from fifth form or lower-sixth form and for males. Males who leave education at this stage are much more likely to enter YTP than to be in employment.

What this means is that Catholics behave differently from Protestants as soon as they leave school. Evidence from the NI LFS suggests that Catholics and Protestants have fundamentally different trajectories through work, education and training between the ages of 16 and 24. The nature of these trajectories needs further analysis. Just from looking at the SELS data, though, it seems that the forces that differentiate Catholics and Protestants continue today and that they operate, for this age group, through differential rates of take-up of education and training rather than through employment and unemployment as in the past.

It is questionable how the religion effect in post-school transitions noted in the SELS analysis can be interpreted. it is no more than a residual once other variables in the model have been accounted for; therefore to attribute this routinely to religious discrimination would be incautious. Other variables, not included in the models, might be important. However, the variables that the analysis controlled for are wide-ranging and it appears that there is some basis for differentiation on the basis of religion.

Why this is so is unclear. It need not necessarily reflect direct discrimination on the basis of religion among young people. Besides the 'discouraged worker' hypothesis of Raffe and Willms (1989), a further hypothesis can be advanced. This is suggested by Teague (1993) and is derived from the way he considers labour markets are socially constructed. The argument is that labour markets are not always open. In areas where there are large employers, an extended internal labour market (EILM) may operate which embeds a company firmly in its locality. Social ties are created between the firm and the community and vacancies are often filled by informal means. These informal job-search methods may well be as important in Northern Ireland f or young people as they seem to be in some parts of Great Britain (Raffe 1985). The evidence from the SELS shows that out of 299 respondents who were in employment, 37.5% said they had found their jobs through 'friends and family.' This response may be, to some extent, a post hoc rationalisation. However, it suggests that questions about the job-search behaviour of young people should be investigated further.

The intention of these mechanisms is not to be discriminatory but rather to regulate a supply of labour. It follows from this that Catholics may be excluded from these social networks, not because they are Catholic, but because the legacy of past social and historical forces. The existence of social networks in Northern Ireland labour markets, and differentials in the way that Protestants and Catholics tie into them, may be one cause of differential take-up of employment and training.

4.8.2 Religious background and qualifications

One finding of the analysis is also worthy of further discussion. It will be noted that the type of qualification gained (passes in Mathematics, Science and English) at GCSE or A Level were rarely significant determinants of post-school destination (and never of the chances of being employed). At GCSE this lack of significance persisted whether a pass was considered to be grade G and higher or grade C and higher.

This finding is significant in light of hypotheses that suggest that a significant proportion of Catholic/Protestant differentials in employment and unemployment opportunities can be explained by the differential human capital endowments of the two groups (Cormack and Osborne 1985; Cormack and Osborne 1989). The basis of this argument is that Catholics lack passes in Science and craft subjects because of the bias toward the humanities of the Catholic school system, and that because these qualifications are more appropriate to employers' needs, they are disadvantaged in the labour market. The conclusion that follows from this is that equalisation of the exam profile of Catholics and Protestants in terms of the number and types of passes ought to lead to a lessening of differentials in employment and unemployment.

To concentrate just on school-leavers who become economically active, this raises questions about how employers perceive and use qualifications. In general, there are two competing interpretations of the value of qualifications in the labour market; the human capital concept and the screening hypothesis. The human capital thesis (Becker 1975) makes a direct link between education and personal economic success. In essence, it is argued that education influences individual productivity and so employability. Differential labour market performance, whether between the qualified and the unqualified, or between Catholics and Protestants, simply results from personal human capital endowment. The second explanation, the screening hypothesis (Gray et al 1983), breaks this explicit link and argues instead that employers use qualifications as a means of ranking potential workers. The clearest distinction between this interpretation, and that of the human capital school, is perhaps in its treatment of qualifications as a proxy for other factors. Qualifications are not important in themselves but rather as a general measure of ability and social/personal attributes. Most skills are learnt on-the-job and qualifications gained at school have little intrinsic relevance. Raffe (1988) argues that employers are not so much looking for specific subject choices but for an indication of exam success that indicates general employability. Hence qualifications have a socially-defined value that can be used to allocate young people to different segments of the labour market (Ashton et al 1990) or different career trajectories (Bates and Riseborough 1993).

This discussion, of course, begs the question about how far employers use qualifications in the recruitment procedure. Ashton et al (1990). using evidence from local case studies of. the British youth labour market, argue that for formal jobs employers do not use qualifications as a formal criterion; their role instead is in sifting school-leavers between broad segments of the labour market. Jones (1985) also points to the importance of qualifications for school-leavers in terms of their impact on job-seeking skills and self perception. The less-qualified, it is argued, have poorer self-images and a lesser career orientation than more qualified school-leavers, that translates into poorer labour market performance.

In Northern Ireland, McWhirter (1989) goes further and suggests that for school-leavers qualifications do not appear to be important in determining whether Catholic or Protestant young people get jobs. On the other hand Miller et al (1993). looking at the fortunes of Northern Ireland graduates, conclude that subject choice explains a large part of the earnings and employment differential between Catholics and Protestants. The evidence for the importance of qualifications in employer selection strategies, and particularly for subject choice, is mixed. It is likely it varies both by type of job and social class; certainly there is no direct and obvious link between subject choice, exam passes and employability as would be hypothesised by some human capital theorists.

To further complicate the picture, geographical variations in labour market conditions also appear to be significant. Labour demand in a locality is dependent on the structure of enterprises in the area in industrial and occupational composition. In Northern Ireland, data from the SELS suggests that the type of employment which young people get is largely unskilled work in construction and retailing. The need for science and mathematics passes to gain access to these jobs is very questionable.

These factors, as well as other environmental variables such as local unemployment rates, have been found to be significant in explaining personal economic success once individual examination attainment and social/personal background have been taken into account (Garner and Raudenbusch 1991; Bagnall 1992).

The discussion has raised two important issues. Firstly, the degree to which the human capital model can be applied to the analysis of differential behaviour by religion in the Northern Ireland youth labour market. The main question concerns the importance of subject choice as a determinant of behaviour over and above the number and level of examination passes as used in other studies. Clearly there are important policy issues arising from the answer to this question. If subject choice does appear to explain inter-religious differences in behaviour, then policies which may lead to comparability between Catholic and state schools (for example the Northern Ireland Common Curriculum following the Northern Ireland Education Reform Order (1989)) might be expected to diminish religious differentials in the labour market in the long-term. Should subject choice appear to be insignificant then there are grounds for the re-evaluation of this line of research. Secondly, given the importance of spatial variations in labour markets, there are strong grounds for investigating the impact of qualifications and social/personal factors on the behaviour of young people in Northern Ireland so as to draw comparisons with Great Britain.

[9] The author wishes to acknowledge the pivotal contribution of Anthony Murphy in developing the analysis presented this chapter.
[10] the model estimates for these, and all later models in the chapter, are shown in Appendix 3

Return to Publication Contents


5.1 Introduction

The purpose of this chapter is to draw together the arguments made about religious differentials in educational attainment and post-school destination, and to give a final overview of the evidence. Three overall objectives, outlined in Chapter One, have guided the report. Firstly, a need to explore the basis of religious inequality among young people given the likelihood that a poor start is a permanent source of disadvantage. Secondly, the aim of providing a further perspective on the debate on religious inequality to be able to make some comments whether religious differentials are a feature of contemporary social and economic forces or a historical relict. Thirdly, interest in the extent to which recent changes in the opportunities open to young people have influenced the balance of religious inequality.

5.2 Overview of findings

Some of the answers are quite clear. Religious differentials exist in educational attainment and post-school destinations. In the case of educational attainment, no significant religion effect remains once other social characteristics have been taken into account. For post-school destinations, after the same social factors have been controlled for, an effect associated with religion remains. This means that the same social variables that 'explain' differential educational attainment fail to explain fully differential fortunes after leaving school.

In terms of the three broad themes of the report some of the conclusions are necessarily more speculative. On the first, the degree to which disadvantage persists, into later life there is no evidence. However, the research already cited (Breen 1991; Narendrenathan and Elias 1993) suggests that this may be case. If this is so, the religious inequalities noted in the SELS in post-school destination could be a potent source of inequality later in life. On the second and third themes it is possible to be more certain.

It appears that religious inequality is the result of on-going social and economic processes. It cannot be said to be an historical relict, reflecting the conditions and opportunities in the society of the 1960s or 1970s as for these school-leavers, who left school in 1990-91, there are differentials in uptake of employment, training and education. This in turn leads on to the third theme of the means by which religious inequality is reproduced. The complication of youth transitions, with extension of education and training opportunities, does not appear to have altered the nature of inequality. Merely it has moved it into different arenas, with the question now not only being who gains work but who gains a training place. and who continues in education.

Some particularly interesting effects are apparent in that it is possible to identify those school-leavers who are most 'at risk' in that religion effects appear to be strongest for them. Looking at the SELS data, Catholic school-leavers who leave from fifth form and lower-sixth form (and especially males) appear to be particularly disadvantaged in the labour market. For upper-sixth form leavers, in contrast, the effect of religion appears to be largely insignificant (though its effects may be manifested in the decision to stay-on at school).

Another feature is that religions, and other social characteristics are associated with complex patterns of multiple social disadvantages. For example, tracing the path of a school-leaver from the time of his/her examinations, it was shown that the labour market status of a pupil's parents was a determinant of success at GCSE. The number of GCSEs gained was then discovered to be a major determinant of gaining a job. But besides this, it was found that school-leavers with both parents employed had a better chance, everything else being equal, of gaining employment This means that a school-leaver from a socially deprived background would not only be more likely to have lower GCSE grades than other pupils, but that this deprived background would have an additional effect in influencing post-school destination thus creating 'double disadvantage'.

In terms of religion these effects might be very strong. It is only in a controlled statistical model that 'everything else is equal'. Given what we know about relative community differentials in SES between Catholics and Protestants, it is very likely that it is disproportionately Catholics who suffer from such multiple disadvantage, Additionally, on top of this, being a Catholic which further increases relative differentials between Catholics and Protestants.

5.3 Policy implications

This finding has policy implications because it suggests that measures to redress religious inequality should be concentrated on this group of school-leavers. Two other findings have policy implications. Firstly, the results that suggest that differential educational attainment can be largely explained by differential SES and possibly by access to grammar school. This implies that policies to address social deprivation, directed through the education system, would also have the result of ensuring equality of educational outcomes. Secondly,, the finding that subject type of examination pass is not a significant determinant of post-school destination f or this group of school-leavers. This suggests that policies to increase Catholic human capital endowment in science/technical subjects will not necessarily alter post-school behaviour. Indeed given the likelihood that most school-leavers, particularly those in the most 'at risk' groups, will probably enter low-skill jobs, the likelihood is that such measures will have no effect. This means that the relationship between educational qualifications and the labour market in Northern Ireland will have to be rethought; at the least these findings suggest that the subject type of exam pass confers no immediate benefit but that any effects will be felt in later life. One mechanism might be that examination passes in science/technical subjects increase 'trainability' which means that the indirect effects of subject type at GCSE and/or A Level will only be apparent after much more time has elapsed since leaving school. Another objection to this line of argument would be that subject type has little bearing on the probability of gaining employment but rather on the type of employment that is gained and that this effect, in terms of earning potential, may also be noticed more in later career.

5.4 Further questions

These questions about the importance of subject type are typical of those raised by the analyses presented in the report. Many of these questions can only be answered by more research and more analysis. In all, three areas of potential enquiry have been identified.

Firstly, there are a number of questions about the effect of qualifications in the labour market, How do employers perceive educational qualifications? Are certain subjects seen as being more relevant in the Northern Ireland youth labour market? Do certain qualifications increase employability and receptiveness to training in later career?

A second theme was the extent to which structural inequalitiesin secondary education contributed to examination success and the growth of religious differentials . In particular, it seems that little is known about how examination policy varies between schools and how this contributes to differences by religion in attainment, The effect of social inequalities in the transfer procedure, as discussed by Sutherland (1993) , may also have religious implications and be worthy of further consideration.

A third set of questions concern the operation of the labour market, If young people find jobs by informal means how effective can fair employment legislation be? Since it appears, from the evidence of other areas, that disadvantage in the labour market is a cumulative process, how effective are later policy interventions given the weak transitions made by many Catholics from school?

These questions cannot be answered by the SELS but are perhaps worthy of further consideration in understanding how religious inequalities arise among school-leavers in the 1990s.


Ashton, D. et al (1990), Restructuring the Labour Market: The Implications for Youth, Macmillan, London

Bagnall, G. (1992). Deprivation and Leaving School, CES, Edinburgh

Bardon, J. (1992), A History of Ulster, Blackstaff, Belfast

Bates, I. and Riseborough, G., (1993). Youth and Inequality, Open University Press, Buckingham

Becker, G (1975), Human Capital: A Theoretical and Empirical Analysis, National Bureau for Economic Research, New York

Boyle, J. (1977), Educational attainment, occupational achievement and religion in Northern Ireland, Economic and Social Review, 8, 79-100

Breen, R. (1991), Education, Employment and Training in the Youth Labour Market, ESRI Paper 152. Dublin

Compton, P. (1991), Employment differentials in Northern Ireland and job discrimination: a critique, in eds. Roche, P, and Barton, B., The Northern Ireland Question: Myth and Reality, Avebury, Aldershot, 40-76

Cormack, R, and Osborne, R. eds. (1983), Religion, Education and Employment: Aspects of equal opportunity in Northern Ireland, Appletree Press, Belfast

Cormack, R. and Osborne, R. (1985), Employment equity in Canada and fair employment in Northern Ireland, British Journal of Canadian Studies, 4(2), 219-231

Cormack, R. and Osborne, R. (1989), Employment and discrimination in Northern Ireland, Policy Studies, 9(3), 219-232

Cormack, R. and Osborne, R. eds. (1991), Discrimination and Public Policy in Northern Ireland, Clarendon Press, Oxford

Cormack, R. and Osborne R. (1993). Higher education in Northern Ireland, in Osborne, R, et al (eds.), After the Reforms: Education and Policy in Northern Ireland, Avebury, Aldershot, 237-251

Daly, P, (1991), How large are secondary school effects in Northern Ireland?, School Effectiveness and School Improvement, 2(4), 305-323

Eversley, D. (1989), Religion and Unemployment in Northern Ireland, Sage, London

Finn, D (1987), Training Without Jobs: New Deals and Broken Promises, Macmillan Education, London

Gallagher, A. (1988), Transfer Pupils at. Sixteen, Northern Ireland Council for Educational Research, Belfast

Garner, C. and Raudenbusch, S., (1991), Neighbourhood effects on educational attainment: a multi-level analysis, Sociology of Education, 64(4). 251-262

Goldstein, H. et al (1993) 9, A multi-level analysis of school examination results, Oxford Review of Education, 19(4). 425-433

Goldstein, H. and Paterson, L. (1991). New statistical methods for analysing social structures: an introduction to multilevel models, British Educational Research Journal, 17(4), 387-393

Gray, J. et al (1983). Reconstructions of Secondary Education: Theory, Myth and Practice since the War, Routledge, Henley

Harbison, J, (1989) (ed.), Growing Up in Northern Ireland, Universities Press, Belfast

Hart, P, (1988). Youth Unemployment in Great Britain, NIESR and Cambridge University Press, Cambridge

Jesson, D. et al (1992), GCSE Performance in Nottinghamshire 1991: Pupil and School Factors, Nottinghamshire County Council, Nottingham

Jesson, D. and Gray, J. (1993), GCSE Performance in Nottinghamshire: Further Perspectives, Nottinghamshire county Council, Nottingham

Jones, P. (1985), Qualifications and labour market outcomes among 16 year old school leavers, British Journal of Guidance and Counselling, 13(3). 275-294

Main, B. and Raffe, D. (1983). Determinants of employment and unemployment among school leavers. Evidence from the 1979 survey of Scottish school leavers, Scottish Journal of Political Economy, 30(1), 1-17

McPherson, A. and Willms, D, (1986). Certification, class conflict, religion and community: a socio-historical explanation of the effectiveness of contemporary schools, in Research in Sociology of Education and Socialization, Volume 6, (ed.) Kerckhoff, A, JAI Press, Greenwich

McWhirter, L, (1989), Longitudinal evidence on the teenage years, in Harbison, J. ed. Growing Up in Northern Ireland, Universities Press, Belfast, 66-92

Miller, R, et al (1993), Higher education and labour market entry: the differing experiences of Northern Irish Protestants and Catholics, in Osborne, R. et al eds. (1993), After the Reforms: Education and Policy in Northern Ireland, Avebury, Aldershot, 255-282

Murphy A, and D, Armstrong (1993), The Incidence and Duration of Male Unemployment in Northern Ireland, Unemployment Forever? The Northern Ireland Economy in Recession and Beyond, Gudgin, G. and G, O'Shea, NIERC, June 1993. 99-113

Murray, D, (1993), Science funding and education in Northern Ireland grammar schools, in Osborne, R, et al eds, (1993). After the Reforms: Education and Policy in Northern Ireland, Avebury, Aldershot, 209-220

Murray, R, and Osborne, R, (1983), Educational qualifications and religious affiliation, in Cormack, R, and Osborne, R, eds, Religion, Education and Employment: Aspects of Equal opportunity in Northern Ireland, Appletree Press, Belfast, 118-145

Narendranathan, W, and Elias, P. (1993), Influences of past history on the incidence of youth unemployment: empirical findings f or the UK, Oxford Bulletin of Economics and Statistics, 55(2), 161-186

Osborne, R, et al (1991), Educational qualifications and the labour market, in Cormack, R, and Osborne, R. eds. (1991), Discrimination and Public Policy in Northern Ireland, Clarendon Press, Oxford, 93-119

Osborne, R, et al eds. (1993), After the Reforms: Education and Policy in Northern Ireland, Avebury, Aldershot

Pahl, R. (1988). Some remarks on informal work, social polarisation and social structure, International Journal of Urban and Regional Research, 12. 247-267

Paterson, L, (1991a), Socio-economic status and educational attainment: a multi-dimensional and multi-level study, Evaluation and Research in Education, 5(3), 97-121

Paterson, L, (1991b). Trends in attainment in Scottish secondary schools, in Schools, Classrooms and Pupils: International Studies of Schooling from a Multilevel Perspective, eds. Raudenbusch, S. and Willms, D.. Academic Press, London, 85-99

Payne, J. (1987). Does unemployment run in families? Some findings from the General Household Survey, Sociology, 21 (2) 199-


Raffe, D. (1984), The transition from school to work and the recession: evidence from the Scottish school-leavers surveys 1977-1983, British Journal of Sociology of Education, 5, 247-265

Raffe, D. (1985). Degrees of informality: methods of job placement among Scottish school-leavers, British Journal of Guidance and Counselling, 13(2). 292-307

Raffe, D. (1988), The story so far: research on education, training and the labour market from the Scottish surveys, in Raffe, D, ed., Education and the Youth Labour Market, The Falmer Press, Brighton, 40-65

Raffe, D, and Willms, D. (1989). Schooling the discouraged worker: local-labour-market effects on educational participation, Sociology, 23(4). 559-581

Regional Trends 1993, CSO, London

Shuttleworth, I, (1993). Pilot Survey of Northern Ireland Secondary Education Leavers: Methodological Report, NIERC, Belfast

Shuttleworth, I. and Armstrong, D, (1993), The characteristics and attitudes of young people on the Youth Training Programme, Training and Employment Agency Bulletin, May, 19-23

Smith, D., (1987), Equality and Inequality in Northern Ireland, PSI occasional paper 39, PSI, London

Smith, D., and Chambers, G. (1991), Inequality in Northern Ireland, Clarendon Press, Oxford

Sutherland, A. (1993), The transfer procedure reformed?, in After the Reforms: Education and Policy in Northern Ireland, eds. Osborne, R. et al. Avebury, Aldershot, 105-126

Teague, P. (1993), Discrimination and fair employment in Northern Ireland, in Teague, P. (ed), The Economy of Northern Ireland: Perspectives for Structural Change, Lawrence and Wishart, London, 141-169

Willms, D, (1992), Pride or prejudice? opportunity structures and the effects of Catholic schools in Scotland, in International Perspectives on Society and Education, Volume 2, JAI Press, Greenwich, 189-213

Whyte, J, (1990), Interpreting Northern Ireland, Blackwell, Oxford

Wilson, J, (1987), Selection for secondary education, in Osborne, R. et al (eds). Education and Policy in Northern Ireland, Policy Research Institute, Belfast, 29-41

Wiseman, S, (1964), Education and Environment, Manchester University Press, Manchester

Return to Publication Contents

© CCRU 1998-1999
site developed by: Martin Melaugh
page last modified:
Back to the top of this page