Stanford or SJSU? Why Skills Now Trump School Name for Bay Area Kids

Standford may be reasonablly priced when it compares with Alpha Schoo but not when comparing with San Jose State.

The interesting part is what happens after graduation. Put a Stanford CS grad and an SJSU CS grad in the same entry-level engineering seat at the same company, and they are paid the same number. Not similar. The same.

The two numbers

What everyone knows: for a family earning above $110,000, Stanford runs about $50,000 a year against SJSU's $22,862. Roughly $110,000–125,000 over four years.

What almost nobody knows: entry-level software engineers in San Jose have a median total compensation of $187,250 — and that number is set by engineering level, not by school. An L3 is an L3. The only variation is at the margins of equity and signing bonus.

One row in that table is blank on purpose. The federal College Scorecard publishes no program-level earnings for SJSU at all — so the Stanford-CS-versus-SJSU-CS comparison that Bay Area parents argue about most confidently is one nobody can actually run.

The number that should worry you more

While parents optimize for admissions, the thing they're optimizing toward has quietly gotten harder to reach.

• Recent grads aged 22–27: 5.7% unemployment, against 4.3% nationally
• Computer science grads: 6.1% — worse than history, communications, or English
• Computer engineering: 7.5%
• Entry-level software engineering roles: down roughly 30% year over year

CS still pays the highest starting salaries of any major. It is no longer the safe bet it was sold as. AI absorbed the junior work — boilerplate code, first drafts, basic test coverage — that used to be the on-ramp.

The degree still works. It just stopped being a guarantee.

What the research says

Decades of it, not a new study:

• Dale & Krueger: college selectivity doesn't predict later income once you account for the student's own ability entering college. Exception: students from low-income backgrounds do gain from elite schools.
• Mountjoy (2025): replicated it.
• Journal of Economic Analysis: field of study predicts earnings more than institution — and selectivity matters more for non-STEM majors than STEM ones.

Not unanimous — some studies find small elite-school premiums. But the summary holds: for engineering, what you study beats where. For finance, consulting, VC, law, and academia, the Stanford network is real and worth paying for.

Which one applies depends on your kid, not on the rankings.

So what do you optimize for instead

If selectivity doesn't move STEM earnings, then a decade spent optimizing for selectivity is a decade on the wrong variable.

The things that do move it — building, debugging, sitting with ambiguity, directing tools instead of being directed by them — aren't built during an application cycle. They're built in ordinary weeks, years earlier.

Four shifts that work at any age:

Praise the process, not the score. "You rewrote your whole approach after the first build collapsed" builds a kid who takes risks. "You're so smart" builds one who protects a record.

Leave the friction in. A puzzle with missing instructions. A broken gadget to diagnose together. A build with deliberately limited materials.

Get three real things built. A student who can explain three projects they designed, broke, and fixed beats a pristine transcript. Robotics or coding at 10 builds what an AP class at 16 can't. Summer is the easiest start.

Audit AI, don't ban it. The work isn't done when the screen says correct — they close the device and explain the mechanism to you. Our tool list and the honest pros and cons.

Three questions for any school tour

Skip the matriculation list. It measures admissions selectivity, not what the school builds.

1. "Show me student work that failed." Polished-only displays train presentation. Visible iteration trains building.
2. "How much of the week has no answer key?" If the answer is "we do a project in May," that's a unit, not a practice.
3. "What replaced the phone time?" Post-AB 3216, every California school reclaimed roughly 40 minutes a day. Some filled it deliberately. Most didn't — and the countries that moved first show the replacement matters more than the ban.

Our Bay Area school reviews flag which schools built this in and which are marketing it.

The bottom line

Stanford opens doors SJSU doesn't, particularly outside engineering. SJSU puts enormous technical talent into Silicon Valley for a fraction of the cost. Both true.

But the entry-level funnel narrowed while the pay bands stayed flat across school tiers. Getting in matters less than it did. Being able to build something matters more.

That capability is assembled over years of ordinary decisions — not in a single April envelope.

Parents Also Ask

Do Stanford and SJSU grads really earn the same at big tech?

‍In the same role at the same level, yes. Companies like Google and Meta set entry-level compensation by engineering level, not by diploma. Median entry-level total comp in San Jose is about $187,250. Equity grants and signing bonuses vary at the margin; the band doesn't.

Is computer science still a safe major?

‍It pays the most and it's harder to enter than it was. CS new-grad unemployment sits near 6.1% — above several humanities majors — while entry-level software roles fell about 30% year over year. Strong paired with demonstrable projects; weaker on its own.

Does school prestige affect earnings?

‍For STEM, research consistently finds little effect once student ability is accounted for. Field of study predicts earnings more reliably than institution. For finance, consulting, law, and academia, selective-school networks carry real weight.