Economic Opportunities, Occupational Class Status and Social Mobility in Northern Ireland: Linked Census Data from the Northern Ireland Longitudinal Study (NILS) John Moriarty, David Wright, Dermot O Reilly and Allen Thurston Centre for Evidence and Social Innovation Centre for Public Health Queen s University Belfast Socioeconomic Status and Employment Both the 28 and 211 Programmes for Government placed economic growth and creation of enhanced high skill labour market opportunities to the fore among the strategic priorities for Northern Ireland. Intertwined with these objectives is an emphasis on the key role of improved educational attainment in driving opportunities for social mobility. The 211 Census enhanced our understanding of the employment landscape in Northern Ireland, which had been shaped by the preceding decades of economic development and by the onset of austerity. Some global trends were reflected in the NI economy, such as the increasing role of retailers as employers, accounting for 18% of the workforce 1. Figure 1: Occupational Class, 1991 and 211 (N=5,376) 1991 211 21.72 47.11 31.17 35.55 25.63 38.81 1 Employer/Manager/Professional 2 Intermediate 3 Routine/Manual/Unemployed 1991- Based on Head of Household of NILS members age 8-17 i 211- Own occupational status in 211, age 28-37 Breaking the population into the simplified scheme of Professional, Intermediate and Routine or Unemployed, the Intermediate category was the smallest in 211. Comparisons with the 1991 Census returns suggest that while these smaller employers may be declining in number, there are growing opportunities for employment at the level of professional and management. 1 Central Statistics Office: http://www.cso.ie/en/media/csoie/releasespublications/documents/population/211/cen211irelandnorthernireland.pdf 1
Social mobility Social mobility is often used as an indicator of societal fairness and openness. The structural changes in the Northern Ireland economy over successive decades suggest that there must be some social mobility in absolute terms. However, as well as overseeing economic change at the macro level, a challenge for policy makers is to break cycles of poverty where they exist for families and communities. Thus, the question arises, to what extent are the opportunities for employment at the various tiers of the economy open to all citizens, and to what extent are people restrained: by their socioeconomic background?; by other factors, such as their gender, level of educational attainment and community background? Addressing these questions requires that we can observe individuals at different points in time and compare employment outcomes for people with different social backgrounds. Linking Census Returns Linked census data provides one window into intergenerational social mobility. The Northern Ireland Longitudinal Study links successive Census returns for a random sample of the NI population. For this study, we identified a cohort of approximately 5, NILS members, whom we can observe at their occupational class "origin" at the 1991 Census, when they were aged 8-17 and again after they have reached working age (28-37) in 211. This design allows for analysis of intergenerational mobility hitherto not undertaken using UK census data. We can directly compare 211 outcomes for people of different socioeconomic backgrounds and in other categories of interest. 2
Findings Figure 2.1: Occupational Class Destination by Origin Full Pop 211 SEC SEC Origin 1 SEC Origin 2 11.23 28.1 17.5 24.32 16.53 9.1 7.27 13.97 9.69 7.89 34.23 15.59 4.94 3.45 6.33 2.99 4.37 31.77 15.7 11.95 5.65 8.7 5.27 4.9 People whose parents were professionals in 1991 were at a clear advantage by 211 over persons from nonprofessional backgrounds. However, it is not clear that there is graded advantage to each successive class. For example, cohort members from intermediate origin classes have a similar profile of outcomes to persons. Figure 2.2: Occupational Class Destination by Origin 211 SEC Origin 3 SEC Origin 4 SEC Origin 5 11.1 9.73 13.12 26.51 21.58 6.56 6.91 13.74 8.54 5.15 23.55 17.83 7.78 8.28 15.47 1.26 7.1 26.12 16.87 13. 7.67 9.7 8.19 5.34 The one group at a noticeable disadvantage to all other groups are those people whose parents were unemployed in 1991. They are more than twice as likely as other groups to be unemployed themselves. Those in employment are more likely to be in lower strata of employment, though a minority do attain intermediate and professional occupations. 3
Overall, the data demonstrate intergenerational stickiness at both ends of the occupational class spectrum and relative fluidity in between. In other words, there is a persistence of effects of both inactivity and poverty, and of affluence. Children of professionals are far more likely to attain professional status early in their careers. If successive generations of families only interact within one class milieu, this may have long-term damage to societal cohesion. This points to a need to prioritise programmes which allow for class mixing, for example, through the education system. Figure 2.3: Occupational Class Destination by Origin SEC Origin 6 Origin 7 Origin 8 1.23 8.99 5.57 15.83 27.27 16.9 1.23 8.1 13.49 24.28 17.86 13.4 6.14 12.64 13.93 8.41 7.53 18.7 14.39 8.38 5.4 1.87 6.18 15.65 A further advantage of having a large representative sample is that we can focus on groups of interest. Some of these have not been studied previously because numbers have been too small to facilitate focus. For others, such as in the case of women, labour market norms have shaped the focus of previous research, which primarily looked at the relationship between men s social status and that of their fathers. 4
Figure 3: Occupational Class Destination by Gender Female Male 9.59 13.1 28.99 22.63 3.81 3.99 17.37 6.13 7.47 19.25 9.92 14.84 1.83 1.27 13.55 8.34 Women are under-represented in higher professional, small employer and lower supervisor occupations, but are over-represented in lower professional, intermediate and semi-routine occupations. While labour market gender parity is not yet reality for this cohort, this does represent a significant increase in workplace participation at all levels for women. Another key contributor to occupational class attainment is education. For our Northern Ireland cohort, there is a pronounced and steady gradient, with each additional level of academic attainment associated with further labour market advantage. Among the higher professions, degree-holders outnumber those with only A-Level qualifications by three to one. 5
Figure 4: Occupational Class Destination by Educational Attainment No Qualifications 1-4 GCSEs 5+ GCSEs A-Levels Degree 1.55 6.76 7.12 12.58 8.48 2.76 2.34 12.77 16.55 12.76 12.1 5.26 22.49 26.14 1.81 1.42 32.39 28.76 29.91 45.65 21.26 21.48 21.97 14.48 7.13 9.52 14.5 7.16 4.12 6.94 4.29 8.99 3.46 4.75 13.4 2.93 3.25 2.43.81 1.61 Furthermore, when we look at factors associated with social mobility, education is more strongly implicated than any other variable. A transition from a routine social origin to a professional is upwards of 7 times more likely for degree holders in comparison with persons who leave school with no educational qualifications. The challenge to policy here is twofold. One is to improve educational outcomes for persons and groups at risk of leaving the academic track prematurely. These groups include those who experience household unemployment and its effects during the early years. The second challenge is around recognition of other modes of learning and knowledge other than academic attainment. In separate analysis, we see that the same returns do not accrue from vocational qualifications, though professional qualifications may have some additional dividend. Given the rise of employment in retail and the prevalence of in-work training and continuous professional development, a system of formal recognition for such training might serve to narrow the gaps accruing from credentialism in the labour market. Finally, an obvious issue in a Northern Ireland context is the historical relationship between community background and employment opportunities. The NI Census suggests that for the cohort of interest, persons aged 28-37 in 211, Catholics were more likely to have come from families with low status occupation or unemployment and less likely to come from professional backgrounds. 6
Figure 5: Occupational Class by Community Background- 1991 Catholic Protestant 2.6 4.46 38.94 26.9 51.41 21.7 1 1 Employer/Manager/Professional 2 2 Intermediate 3 3 Routine/Manual/Unemployed However, it appears from Figure 6 that there were few differences in by community background by 211. There remains a slightly greater proportion of unemployed among persons whose families identified as Catholics in 1991. However, overall, the experience of structural inequalities for Catholics in this cohort is greatly diminished. This necessarily implies that Catholics in this cohort have experienced greater intergenerational mobility when we compare with social origin. However, the effect of this mobility has not been to create any labour market advantage per se for Catholics, but to reduce previous inequalities. 7
Figure 6: Occupational Class by Community Background- 211 Catholic Protestant 11.19 11.42 24.41 16.6 9.55 6.55 13.13 9.68 9.44 24.43 16.97 8.8 7.94 14.74 9.58 6.13 Conclusions There are several advantages to using linked Census data to explore longitudinal trends in this area: The study cohort is randomly chosen from the population and, hence, is representative The study is not skewed by selective participation and dropout The baseline data are not reliant on a respondent s recall of their parents occupation and social position. We do find evidence of cycles of intergenerational privilege and poverty, though there has also been some mobility. Children of professionals are far more likely to attain professional status early in their careers. This may affect societal cohesion, if successive generations of families experience affluence without experience of deprivation. Similarly, intergenerational unemployment is likely to have an alienating effect. Education is the strongest predictor of occupational status for the generation we are studying. However, most of the returns are to academic and third level qualifications. Given that training is available across the lifecourse it may be profitable to examine how to better reward other forms of training and knowledge. Overall, we find recent differences by community background to be very small in comparison with differences by gender, education and social class origin. We suggest that it is in the public interest to move the debate beyond a focus on religion to an awareness of broader socioeconomic structures and inequalities. i 1991 Occupational Class is calculated based on a retroactive coding of occupation and establishment size. The three-class scheme is more conceptually consistent with NS-SEC than the eight-class scheme, therefore we use the three-class version when comparing class distributions directly. 8