The Value of College – Goldman report
Get The Timeless Reading eBook in PDF
Get the entire 10-part series on Timeless Reading in PDF. Save it to your desktop, read it on your tablet, or email to your colleagues.
What if I Told You …College May not Be Worth it?
The average return on going to college is falling. For the typical student the number of years to break even on the cost of college has grown from 8 years in 2010 to 9 years today. If current cost and wage growth trends persist then students starting college in 2030/2050 will have to wait 11/15 years post college to break even. 18 year olds starting college in 2030 with no scholarship or grants will only start making a positive return when they turn 37.
What if I told you college may not be worth it. These are averages though. The distribution of wages is changing though – using SAT scores as a proxy, the gap in alumni earnings between colleges in the 99th percentile of scores and the 99.9th is as big as the gap between the 1st percentile and the 20th. If future wages for an individual joining college today were 30% below the average wages expected for college graduates, they wouldn’t break even until they were over 50. Graduates studying lower paying majors such as Arts, Education and Psychology face the highest risk of a negative return. For them, college may not increasingly be worth it.
Why It Matters? The choice of college and major are more important than ever to students given the changing return profile. (US student debt is now more than $1tn). The still high levels of skilled vacancies shown in Exhibit 10 despite record numbers of undergraduates points to a demand and supply mismatch (linked to this is the shortages in trade jobs; skilled trade vacancies screened as the most difficult job to fill in Manpower’s annual survey for the sixth consecutive year in 2015). Corporates may have to do more themselves and develop their own talent identification systems. New entrants and business models are emerging to meet some of these challenges.
Studying the cost of studying
We calculate the economic return on college education as the total all-in cost of college (net of grants and scholarships) and the wages foregone during the 4 years of study versus the wage premium that undergraduate degree holders enjoy versus high school graduates over their working life. Since 2009, college fees have risen by 10.6% in real terms, versus a 0.9% and -0.1% change in high school and college graduates’ wages. This calculation is just the economic return and of course there are other broader factors which matter greatly.
For whom the bell-curve tolls
Choices matter – median wages of recent college graduates in architecture / engineering are 51% higher while wages of Arts graduates are 15% lower than the median across majors. A recent Economist study of US colleges based on US Department of Education data shows that the median annual earnings of MIT graduates ten years post college is $91,600 (almost twice the national median earnings). At the other end of the scale graduates from the bottom 25% colleges earn less, on average, than high school graduates. Another way to think about the changing shape of the labor market is that graduates of colleges with a 90th percentile SAT score entrance requirement make $11,700 more per year than those from universities in the 10th percentile.
Skills sought versus skills taught
With more graduates than ever, it is an interesting paradox that skilled vacancies in many industries are high. Take the example of the IT industry; as evident in Exhibit 10, the sector has a high and growing vacancy rate (3.6% since 2011, up from 2.8% between 2001- 05) but less than 3% of degrees conferred are in computer science (vs. 6% in education and 10% in social sciences and history). More people graduating has blunted the signaling power of an undergraduate degree contributing to a rise in the number of masters degrees (now 7.6% of population versus 5.4% in 2000) which adds to the cost of education. But the vacancy rate does imply that the increased need for specialization is a factor and that the labor market is moving faster than education can respond. Arguing against this is the threat from artificial intelligence to pattern recognition and human decision making jobs (47% of total US employment is at risk from computerization according to a recent Oxford University study). This hollowing out of the labor market risk makes future-proofing very difficult, but there is a clear growth opportunity in professional training.
What could change?
Two things in particular stand out as potentially disruptive to universities. First if employers changed their attitude toward nontraditional sources of degree awards. Massive open online courses (MOOCs) are the most obvious threat (30% of undergraduates already take some classes on line), but it could also be companies creating their own de facto degrees. Udacity for example offers nanodegree programs where curricula are designed in partnership with companies like Google, AT&T, Facebook, Salesforce and Cloudera. Second, for a broader new system of signaling and talent identification, look again to the tech sector; an increasing numbers of companies are using GitHub (a software development tool used for writing, storing and collaborating on code) to view coders’ portfolios of work as a better talent indicator than their academic resume. Consulting firms EY and PWC have both said they will use their own testing systems for recruitment rather than relying on academic grades.
See the full PDF here.
Alternatives to College
Joel Greenblatt on Valuing Stocks
Article by CSInvesting