# Degree signaling erodes by sector, then stratifies by institution (2025–2045)

**Claim:** The bachelor's degree loses signaling value first in fields where ability is demonstrable via machine-evaluable work samples — software, data, design, digital marketing — because AI-based skill assessment replaces the degree's screening function. This is already visible: tech postings dropping degree requirements, and skills-first hiring platforms scaling through 2030–2035.

**The counterintuitive part:** signaling value doesn't just disappear; it *migrates upward* to institutional selectivity and brand. When "has a degree" stops discriminating among candidates, "which degree" becomes the signal. Consequences:

1. **Total enrollment does not collapse** — the degree shifts from a labor-market signal to a stratification marker and a default credential for regulated/licensed professions (law, medicine, engineering sign-off, academia), which retain it.
2. **The mid-tier loses most:** comprehensive regional universities and mid-selectivity state schools sit where signaling decay hits hardest — expensive enough to sting, not prestigious enough to signal.
3. **Regional pattern:** the decline concentrates where large, mid-selectivity public flagships anchor the state's higher-ed market (Sun Belt growth states included — the SEC-region enrollment boom is a volume story, not a prestige story). Coastal elite privates and selective publics (Michigan, UNC, UVA) retain signal premium. Some formerly prestigious privates in weak regional labor markets face the squeeze from both sides.

**Falsifiable check:** if by 2032 postings data shows degree requirements stabilizing or rising in tech, and wage premia for mid-tier grads hold constant vs. elite grads, this claim is wrong.