The most expensive degree is the one that looks socially useful while quietly turning your twenties into a repayment plan. A lot of families still treat university like a single product, as if every four-year qualification lands in the same place. It does not. Some degrees buy access. Some buy time. A few buy money. The rest buy a framed photo, a graduation party, and a slow argument with the bank.
Computer science is in the third category, and by a long way. Recent ROI breakdowns put it at around 2,900 percent over a lifetime, with nearly $900,000 in pure profit inside the first 10 years for the right graduate. That kind of number changes the conversation fast. Once you compare the money, a lot of the “study anything as long as you get a degree” advice starts sounding like something said by people who have never had to calculate interest on student debt.
The degree market is not equal
The polite lie is that a degree is a degree. The messy truth is that some qualifications are financial assets and some are expensive hobbies with lecture notes.
Georgetown University’s College Payoff data has been saying this in less dramatic language for years. STEM degrees produce much stronger lifetime earnings than humanities degrees, and computer science sits near the top of the pile. The Federal Reserve Bank of New York has also shown the same pattern in a different form: computer science, engineering and health tech graduates tend to face lower unemployment and stronger wages than graduates from fine arts, humanities and education.
That gap matters more now because debt is not abstract anymore. A public university degree in South Africa can easily run from about R40,000 to R70,000 a year in tuition alone, before transport, rent, meals and data start chewing at your budget. Over three or four years, you can land anywhere from roughly R120,000 to more than R280,000 in total costs. If you borrow, that bill does not evaporate because the graduation photos looked good.
Engineering, finance and tech are still the cleanest high-return categories. If you want a blunt answer, those are the fields with leverage. The rest of the market is full of degrees that sound noble and pay like a long apology.
Why computer science keeps winning
Computer science does not win because it sounds smarter. It wins because one person with the right skills can build, ship, automate and sell things that used to require a department.
A junior software developer in South Africa can start somewhere around R25,000 to R45,000 a month, which is roughly R300,000 to R540,000 a year. Once experience kicks in, R70,000 to R100,000 a month is not fantasy. Compare that with many graduates in humanities or other non-technical fields, where starting salaries often sit closer to R10,000 to R20,000 a month. That is not a small difference. That is the difference between paying debt and letting debt become your personality.
The global 2,900 percent lifetime ROI figure gets attention because it is so aggressive, but the logic behind it is easy to see. A CS graduate can reach a much higher income band early, move into better-paying specialisations, and keep compounding from there. The “nearly $900,000 in pure profit within 10 years” claim sounds absurd until you remember that the market pays for output, not just attendance.
Then there is resilience. A useful degree is not only about your first salary. It is about what happens when the economy gets weird, when hiring slows, when the fancy company freezes roles. Computer science still opens doors because the work is directly tied to systems businesses cannot ignore. Software breaks. Data piles up. Security gets attacked. Cloud bills need trimming. Someone has to fix the mess.
What costs the least and pays the most
The most obvious trap is assuming “tech” means “four years at university or nothing.” That is lazy thinking. There are cheaper routes, faster routes, and in some cases better routes if your goal is income rather than ceremony.
A bootcamp can be enough if you take it seriously
WeThinkCode_, HyperionDev and CodeSpace Academy all offer shorter, intensive training in programming and software development. Expect roughly 3 to 12 months depending on the programme. Entry usually starts with an aptitude test or application process, not a matric subject checklist. Your first step is simple, apply, finish the selection process, and build projects like your life depends on it, because the portfolio is the real CV.
Online learning is not fake learning if you finish the work
Coursera, edX, freeCodeCamp and Udacity can get you from zero to job-ready basics without forcing you into a residence with bad coffee and worse Wi-Fi. Time ranges from a few weeks to several months depending on the course. Entry requirement is mostly discipline and a laptop that can survive a browser with 47 tabs open. Start with one track, finish one project, then post it on GitHub before you start collecting badges like they are Pokémon.
Certifications can shortcut the gatekeepers
AWS Certified Cloud Practitioner, Microsoft Azure certifications, CompTIA Security+ and the Google IT Support Professional Certificate all speak loudly in hiring conversations. They usually take weeks or a few months, not years. Entry requirement is usually basic computer literacy and the willingness to study like an adult. First step, pick one vendor, learn the syllabus, and practice until the exam no longer feels like a foreign language.
TVET and apprenticeships are underrated
Some TVET colleges now offer IT support, software and network-administration paths at a lower cost than a full university route. Apprenticeships and internships can also be the fastest bridge into actual work. Duration varies from a few months to a few years. Entry is usually more flexible than university, but the first step is the same everywhere, apply early and show evidence that you can solve problems, not just pass them.
How to check this is legit
Any funded or fee-charging programme should be treated like a landlord with a nice smile. Nice presentation is not proof.
Check whether the programme has clear outcomes, named employers, or graduates you can actually verify on LinkedIn. Ask what projects you will leave with. Ask who teaches, how often the curriculum changes, and whether the certificate means anything outside the marketing page. If a bootcamp cannot explain where its graduates work, that is a red flag wearing a hoodie.
For funded offers, read the repayment terms before you get excited. If the model is income-share based, calculate the worst case, not the brochure version. If the programme promises a job, ask for the placement rate and the date it was measured. If the answer is vague, the offer is vague.
The other quantitative degrees people keep sleeping on
Computer science gets the loudest praise, but it is not the only sensible bet.
Actuarial science is brutal, mathematical and expensive in brain cells, which is exactly why it pays. Data science and analytics are still pulling strong demand because every company is drowning in numbers and half of them do not know what the numbers are saying. Quantitative economics, econometrics and operations research are all useful if you want to work in finance, consulting, logistics or policy. Biostatistics and bioinformatics matter more than most school leavers realise, especially as health tech and pharmaceutical research expand. Applied mathematics and computational maths are the sort of degrees that quietly end up in finance, cryptography, modelling and systems work.
These are not glamorous qualifications. They are money degrees. There is a difference.
Where the real demand sits right now
If you want to be practical, aim at the parts of tech that businesses cannot delay. AI and machine learning, cybersecurity, cloud computing, data engineering and DevOps are all sitting high on the demand stack. E-commerce, logistics, solar installation, battery systems and data centres are also creating work that did not matter much 10 years ago and matters a lot now.
A young South African does not need to wait for a perfect local job listing to appear either. Upwork, Fiverr, Contra, Shopify, GitHub, YouTube, TikTok and Etsy are all live channels for earning, building proof, or selling directly. If you can code, design, edit, automate or analyse, you can get paid in dollars, euros or pounds through PayPal, Payoneer or Wise. That exchange rate is not a footnote. It is the point.
Load shedding and data costs still exist, which means the person who learns how to work around them has an edge. A cheap laptop, a stable power plan, and enough connectivity to ship projects are more valuable than another round of career advice from someone who last used a campus computer lab in 2009.
So is college still worth it
Yes, if the degree is tied to a market that pays.
No, if you are paying university prices for a qualification that leaves you fighting for low-wage jobs while debt grows teeth. The market has already decided that some majors are worth the money and some are mostly expensive identity exercises. Pretending otherwise is just sentiment dressed up as guidance.
If you are choosing now, choose like someone who plans to be employed, not applauded.
Computer science is not the only route, but it is one of the clearest ones. If you want to argue that a social experience justifies a degree that may take decades to pay off, bring the numbers with you. If you want to study something that gives you leverage, computer science keeps making the case for itself.
Drop your major. The argument starts there.
