Don’t study your high school sweetheart

I have always found the way society forces young people to specialise so early to be slightly ridiculous. Barely old enough to vote or purchase alcohol, young people need to make a choice which will inevitably affect the rest of their lives. They don’t really know what is out there, remain heavily influenced by the small sample of those around them and false perceptions of choice outcomes. I wrote this essay on the back of a small research study conducted at university, and whilst it is more formal than my usual blogs – I think the results are very thought provoking.

1. The Dilemma

Australian universities offer short, specialised degrees. Admissions typically require prospective students to select a discipline before commencing study, which involves selecting from 100+ degrees before stepping foot on campus. Some universities require a further sub-specialisation (major) choice. Operating under high levels of uncertainty, students make initial choices with low discipline$^1$ match quality. Uncertainty leads to utilisation of the availability heuristic (amongst others) resulting in a bias for disciplines studied during school. Sub-optimal matches are preserved by a reference level of faster graduation times and high switching costs.

By 2050, 45% of each age cohort of Australians will complete undergraduate degrees [1]. Government and private spending total AU15.7B p.a. on domestic undergraduate education [2].$^2$ Whilst graduates on average outperform those without degrees$^3$, Figure 1 demonstrates the significance of discipline selection for the financial return of a degree. For governments, tax revenues, labour force participation and unemployment are closely linked to discipline selection: For workers, education-occupation match quality is correlated to wages and job satisfaction. Vieira (2005) and Hersh (1991) found that perceived over-education$^4$ has significant$^5$ adverse effects on job satisfaction [3, 4]. Allen & van der Velden (2001) find feeling over-educated has strong negative effects on wages, whilst skill mismatches adversely affect job satisfaction [5]. 2. The Research 2.1 Sub-optimal choices under uncertainty Most undergraduates enrol with school experiences fresh in their minds; 72% of a typical cohort is 19 or younger, and 45% left school three months prior [1]. In a survey conducted with 206 domestic graduates and current students, only 29% indicated they were “very certain” about their initial degree choice. 50% of participants either; (1) changed their discipline (47%) or (2) discontinued their studies (3%).$^6$ In reflection, 43% said they chose the wrong degree and 7% said they chose the right degree but the wrong major. Kahneman and Tversky (1982) suggest under conditions of uncertainty, students may utilise availability as one heuristic to aid their decision making [6]. Required to choose a discipline with uncertainty about match quality, students may more easily retrieve instances of themselves having a match to disciplines studied in school compared to others.$^7$ Comparing initial enrolment disciplines to disciplines which students “change into”, we find evidence students tend to “over-enrol” into school disciplines. As Figure 3 shows, there is a net change out of school disciplines (-4%) and a net change into other disciplines (+4%) suggesting this availability effect fades over time. The result is not significant (p=0.25)$^8$ but would be for similar proportions at larger sample sizes. Students who enrol into school disciplines are also significantly (p=0.0008)$^9$ more likely to change their degrees (Figure 4), suggesting more uncertain students tend to initially enrol in school disciplines and later change. 2.1 Changing disciplines: harder than it may seem Sub-optimal discipline matches are preserved by the reference level of a "change-free" graduation time. The structure of Australian degrees means students are unable to make even one discipline change without “adding time” to their degree.$^{10}$ Kahneman and Tversky’s reference-dependent model of choice suggests this perceived loss in time will have a substantial impact on preferences [7]. The initial endowment of post-university time is overvalued, whilst the gains from changing disciplines are undervalued. Our survey data aligns with this theory; 38% of the students who did not change their discipline (but considered it) highlighted the loss of time as a reason. The second most common reason, uncertainty about a new choice, highlights perceived high switching costs – an indirect measure of this loss. 3. The remedy Student-discipline match quality can be improved by delaying specialisation and encouraging experimentation through a reduction in the associated perceived losses. 3.1 Delayed specialisation A study by Malamud (2011) compared post-university behaviour between graduates specialising before commencing (England) and specialising after 2 years (Scotland). Malamud found Scottish graduates are 10-20%$^{11}$ less likely to switch to an unrelated field in later careers, indicating higher discipline match quality in Scotland [8]. Scottish graduates also choose disciplines not available in school in higher proportions, reducing the impact of our hypothesised availability effect. Engineering and law$^{12}$ are chosen in higher proportions in Scotland, whilst math, physical sciences and art is chosen in higher proportions in England (Figure 6).$^{13}$ The usual justification for early specialisation prioritises the value of job specific skills. Malamud’s study contradicts this assertion, concluding that in this case “the benefits to increased match quality … outweigh the greater loss in skills”[8] For Australian universities, this implies absolving two-tier specialisation (degree and major) into a singular specialty, and initial enrolment into a generic degree.$^{14}$ Given universities with earlier specialisation may appear more attractive to prospective students, we suggest government intervention to coordinate a simultaneous transition [9]. 3.2 Encouraging experimentation Delayed specialisation should be accompanied by a reduction in the perceived cost of experimentation. We suggest the following options: 1. for $n-year$ degrees the longest pre-requisite chain$^{15}$ is reduced to $n-1$ years, or, 2. degree lengths are extended by 1 year to include more space for experimentation Longer degrees would have minimal impacts on completion rates; currently only 34% of students complete their degrees within the theoretical length [10]. Experimentation could also be encouraged by requiring students to complete foundational subjects for at least 2-3 disciplines in their first year with sufficient differentiation. Directions towards students should explicitly encourage experimentation, in contrast to many university websites which advise choosing a major as early as possible. Occupational choice theory suggests if the quality of job matches is difficult to forecast, turnover will be higher [11]. In marriage models, cohabitation has been shown as primarily a means of experimentation with potential partners [12]. It is perhaps surprising in so many other areas of life, experimentation is feverishly encouraged, but information poor university students are not given such freedoms which will make them better off. 4. Conclusion Improved discipline matches will lead to improved welfare for graduates in the form of higher productivity in the workforce, higher incomes, and increased job satisfaction. Reducing the availability effect could also increase specialisation in areas such as engineering and teaching, easing skill shortages. As is usually the case, the decision context of university choice is not neutral. Given the direction of bias is potentially sub-optimal, universities and the government can improve societal welfare by making it easier for students to make better choices [13]. Whilst students primarily gain from this change, universities can also gain significantly from accepting their responsibility as match quality testing grounds; well matched graduates will be more employable and more successful researchers. Footnotes 1. Since Australian universities offer specialisation at both the degree and major level, discipline will be used as a grouping term to refer to the field of study (which can either be a degree, major or combination of both) 2. See Appendix for a breakdown of costs. Calculated from Grattan Institute - 2018 3.A graduate has median lifetime additional earnings of585K (female) - \$789K (male) compared to a regular year 12 student, and 1/3 the average unemployment rate [2, p. 80].

4. Over education commonly occurs after switching to an occupation unrelated to education, implying study was unnecessary or in the wrong discipline

5. 1% significance level

6. The low discontinue rate compared to the overall figure (43%) is likely due to the sample consisting of more current students.

7. See appendix for the approach to classification of secondary school vs other disciplines

8. Two sampled Z test of proportions (unpooled)

9. Using McNemar’s test

10. Unless they have coincidentally done a double major or double degree from the start which may explain their rising popularity

11. For two different data sets

12. In our study, law is a school discipline. Law is not offered at all during school in Scotland

13. GCE A-Levels offer specialised subjects, however, this does not detract that most students would complete mathematics, physical sciences, and arts in school.

14. For example, the US system only offers a BA and a BS at degree level.

15. A pre-requisite chain occurs when a student needs to do course 1A in order to do course 1B and course 1B to do 2A and so forth for their entire degree.

References

• [1] DESE, “Undergraduate Applications, Offers and Acceptances, 2020,” Australian Government, 2020.
• [2] Grattan Institute, “Mapping Australian Higher Education,” https://grattan.edu.au/report/mapping-australian-higher-education-2018/, 2018.
• [3] J. Hersch, “Education Match and Job Match,” The Review of Economics and Statistics, vol. 73, no. 1, p. 140144, 1991.
• [4] J. A. C. Vieira, “Skill mismatches and job satisfaction,” Economics Letters, vol. 89, no. 1, pp. 39-47, 2005.
• [5] J. Allen and R. van der Velden, “Educational mismatches versus skill mismatches: effects on wages, job satisfaction, and on-the-job-search,” Oxford Economic Papers, vol. 3, no. 1, pp. 434-452, 2001.
• [6] A. Tversky, D. Kahneman and P. Slovic, Judgement under uncertainty: Heuristics and Biases, Cambridge: Cambridge University Press, 1982.
• [7] A. Tversky and D. Kahneman, “Loss Aversion in Riskless Choice: A Reference-Dependent Model,” The Quarterly Journal of Economics, vol. 106, no. 4, pp. 1039-1061, 1991.
• [8] O. Malamud, “Discovering One's Talent: Learning from academic specialisation,” INDUSTRIAL AND LABOR RELATIONS REVIEW, vol. 64, no. 2, pp. 375-405, 2011.
• [9] P. B. Gupta, P. M. Saunders and J. Smith, “Traditional Master of Business Administration (MBA) Versus the MBA With Specialization: A Disconnection Between What Business Schools Offer and What Employers Seek,” Journal of Education for Business, vol. 82, no. 6, pp. https://doi.org/10.3200/JOEB.82.6.307-312, 2010.
• [10] OECD, “Education at a glance - 2019 - Australia,” 2019. [Online]. Available: https://www.oecd.org/education/education-at-a-glance/EAG2019_CN_AUS.pdf.
• [11] R. A. Miller, “Job Matching and Occupational Choice,” Journal of Political Economy, vol. 92, no. 6, pp. 1086-1120, 1984.
• [12] M. J. Brien, L. A. Lee and S. Stern, “COHABITATION, MARRIAGE, AND DIVORCE IN A MODEL OF MATCH QUALITY,” International Economic Review, vol. 47, no. 2, pp. 451-494, 2006.
• [13] R. H. Thaler and C. R. Sunstein, Nudge: The Final Edition, Penguin, 2021.
• [14] ABS, “Qualifications and work,” 2019. [Online]. Available: https://www.abs.gov.au/statistics/people/education/qualifications-and-work/2018-19.
• [15] Department of Education, Skills and Employment , “Completion Rates of Higher Education Students - Cohort Analysis, 2005-2019.,” 2020. [Online]. Available: https://www.dese.gov.au/higher-education-statistics/resources/completion-rates-higher-education-students-cohort-analysis-2005-2019.

Appendix

Other data

• "I changed majors within the same degree. And simply because the original major i chose wasnt as interesting as uni (but was something i enjoyed in high school)"
• "It would have been very useful to do one finance class in the first semester to prevent stuffing up degree progression had I changed majors"
• "I realised I made the wrong choice, but stuck it out and graduated with honours anyway. Now I'm going back to do a Masters of Teaching and go down that career instead"
• "I came into university liking maths and commerce in high school and came into uni in a commerce (finance) and advanced mathematics (statistics) degree. I realised a year or 2 in that it was not for me"
• "I think this is a super important issue that needs to be solved. However, despite wasting a little bit of time, I do appreciate the learnings I had along the way as I tried to figure out what I liked."
• "School maths isn't the same as the maths stuff you do for various degrees. I was terrible with maths at school for a variety of reasons but I was also 16/17 and distracted with other things. School learning experiences had too much impact on what I chose to do, I wish I knew what commerce would have been like rather than doing really in-depth maths equations."
• "Started studying after high school in 2013, enrolled in an arts degree to study english lit because that's what I was good at in high school. Hated it, dropped out."
• "I'm here at ANU and it's honestly amazing at how much flexibility I get in my double arts/busn degree. I would've been extremely unhappy if I went to Melbourne or Sydney as I've heard the degrees are way more rigid"
• "I would have preferred to have spent some time trying out many different subjects so that I could get a general feel/affinity for tertiary-level studies, because how you are taught and what you study in university is completely different to high school."
• "It's hard cause you want to explore options but run out of space for electives later on"
• "Getting work experience would really help as far as making decisions goes. A lack of awareness and belief in fictional series gets students to make decisions based on ignorance. Maybe going to uni 2-3 years after high-school would be best. Kids need to mature a little."
• "Universities actually make it very difficult to transfer courses and credit points, which should be considered - it probably deters people from changing."
• "I think part of the issue is that you don't know what you don't know. It's extremely difficult for high school students to understand what different degrees actually entail."