Gender‐specific peer groups and choice at 16

Abstract

The purpose of this paper is to examine the influence of gender‐specific peer groups on students’ intentions and realisations to stay on into post‐compulsory education at the age of 16. This is an important area for research as one of the UK government’s aims is to achieve a 50% staying on rate in higher education at the age of 16. However, this might not be achievable because it is demand‐constrained: not all students want to stay on in education at 16. Peer groups are known to affect demand decision‐making and such affects are also known to be stronger for boys than for girls. Here we explicitly model the demand for post‐compulsory education as a function of peer groups. Our results suggest (1) boys’ intentions and realisations are influenced mainly by their male peers, and (2) girls’ intentions are influenced by their whole peer group while their female peer group influences their realisations. Policy targeted to increase participation rates should recognise these gender differences.

Keywords:

• Education economics
• School choice

Acknowledgements

We are grateful to all of the schools that participated in the original study and the young people who completed the questionnaires, the Bradford Training Enterprise Council for funding and support for data collection for the original report. Thanks also go to Paul Dunne, Peter Howells, Rob Cornes and Wayne Thomas for their helpful comments on earlier drafts of this script. None of the above bear any responsibility for errors or conclusions expressed in this paper.

Notes

1. This endogenous relationship will mean results generated from ordinary least squares estimation procedures will be inappropriate. Evans et al. (1992) employ ordinary least squares and simultaneous equation estimates to identify the importance of peer groups on teenage school dropout behaviour and pregnancy. Their results suggest that single equation estimates that do not control for endogeneity often indicate significant peer group effects that vanish with the use of simultaneous equations. Data availability problems often preclude such estimates. Evans et al. (1992) do not reject the role of peer groups out of hand, but rather they find that the effects are to some extent under the control of the individuals who choose to become members of the peer group.

2. The role of peer groups is likely to be prevalent across all spectra of activity. Some research that would appear to corroborate such findings has been conducted into offending among young people. Jamieson et al. (1999) found a higher rate and frequency of offending among boys than among girls. Girls were more likely than boys to report having never offended and were less likely than boys to be classified as persistent offenders. At least some of this variation could be attributed to different peer influences. The authors found that peer pressure was considered by young people to be particularly influential in respect of offending among boys.

3. Francis (2000) supports this with evidence that girls are now considering a far wider range of occupations than they did in the 1980s and that these occupations increasingly require degrees as an entry requirement (such as doctors or solicitors).

4. Benenson et al. (2002) examine whether females exhibit greater discomfort than males in competitive contexts, which could account for females’ greater avoidance of direct competition. They found that female children exhibit significantly more discomfort than males when they were asked to choose one leader for their group, even though there were no gender differences in the level of participation or length of negotiations during the process of leadership selection. They provide evidence that females exhibit higher levels of discomfort than males when competing directly with same‐sex peers. One plausible proximate reason for sex differences in engagement in direct competition is females’ more negative emotional response.

5. Peer group effects may well have an effect through the psychological consequences of success or failure. Okamoto (1999) investigates the relation between the existence of intimate others and fear of success for 302 undergraduate students and shows that males tend to be afraid of success when they have fear of failure, while females tend to be afraid of success when they think that success would violate their traditional feminine role. In an educational setting, this might transpire to have an affect on the decision to stay on into post‐compulsory education: if boys’ peers are staying on into post‐compulsory education then boys might feel they are being perceived as a failure if they decide not to stay on; if girls peers increasingly perceive educational success to abide with the traditional feminine role, and increasingly do so with the societal change from production‐led to information‐led output, then staying on in education will strengthen their perceptions that they are conforming to the expectations placed on them by society and not to compete against it.

6. Various data sets indicate that students frequently have unrealistic expectations of their future wage streams.

7. The data set indicates which individuals study in each school; it does not indicate whether individuals are friends of other students and whether they are considered their immediate peers by the student. This leads us to have to assume that the whole cohort represents the student’s peers. Nevertheless, this might be a correct assumption because direct peers are often comprised of different people for each person: different groups of the cohort will influence different people but each person will then have an effect (to varying extents) on each of the other students.

8. We assume that the school‐specific component in the error term is uncorrelated with the peer group composition.

9. Data were obtained several months before the pupils sat their GCSE exams.

10. Data were obtained several months after the pupils sat their GCSE exams.

11. These peer group effects might well be illustrative of some type of school culture effect that is not captured by the fixed effects in the regression.

12. This peer group composition is likely to be correlated with current and post determinants of the staying on decision through systematic choice of neighbourhood and school by families.

13. Selective schools might promote academic achievement that fosters the academic culture and encourages students to stay on. Jesson et al. (1991), Cheng (1995) and Mortimore et al. (1997) all found that the pupil’s school type did in part explain post‐16 destinations and Smith & Tomlinson (1989) found that the school attended influenced exam results at age 16 and, as a consequence, the decision to stay on. However, including a dummy variable to capture school type did not prove to be statistically significant or alter the results in any important manner in this study.