The current facet is: Regional variation (SEC growth vs. northeast/rust belt decline).

The session substance (the increment) is at the bottom of the last block, the "Claim (facet: regional variation...)" text:

- By early 2040s, US higher ed splits into two geographic regimes.
- SEC/Sun Belt states (TX, FL, GA, TN, AL, Carolinas) grow on all axes: public flagships add estimated 20–35% enrollment 2025–2045 on domestic migration, employer relocation, state investment in AI/compute infrastructure (data centers, corporate AI campuses) pulling faculty, federal R&D dollars, industry partnerships south.
- Northeast/Rust Belt shrink on three axes at once: 25–35% fall in college-age population, outmigration of AI-linked employer and tax base, consolidation — small privates and regional publics in PA, OH, NY, MI, New England close/merge/convert at 3–5x national per-capita rate, totaling roughly one in four four-year institutions in that belt by 2045.
- AI is an accelerant of the existing migration pattern, not an independent cause — compute-hungry industry follows cheap land, power, low taxes to the South; university prestige follows industry funding.
- Endgame: regional brand monopoly — Southern flagships become default national universities; Northeast retains only its elite tier; displaced students migrate south (compounding bifurcation) or exit degree-seeking entirely.

Now — prior filed notes relevant to this facet (from the summaries in Note 1, Note 7, Note 8):

- Consolidation note: demography (births 4.32M → 3.59M, ~17% decline, CDC/NCHS; WICHE HS grads peak ~3.9M ~2025; Northeast/Midwest/Plains lose 10–20% college-age per WICHE; TX/FL/TN/GA/Carolinas grow; Census domestic migration to Sun Belt). Program mix × region: SEC/Sunbelt publics +10–25% seats through ~2040 as legislatures fund as economic development; northeast small privates −30–50% by 2040; "program mix picks the winners, regional demography picks which losers can still switch"; three stacked currents: demographic cliff, migration to Sun Belt, AI disruption landing on white-collar task clusters in legacy NE/Rust Belt metros; flagships everywhere partially insulated.
- Research facet: geography follows research dollars, not football geography; SEC growth state-policy-driven, not conference/research-endogenous.
- Cost structure: balance-sheet dependence of timing over region — "regional patterns are mostly proxies for endowment wealth, tax base, cost of attendance."
- Enrollment demand: Sun Belt/SEC flagships grow on population inflow even as credential value falls; Southern enrollment growth masks rather than refutes credential-value erosion.

So what is genuinely NEW in this session's increment for the regional facet?

1. The three-axis simultaneous shrink mechanism for NE/Rust Belt: demographic decline + AI-linked employer/tax base outmigration + consolidation. Prior notes had demography and AI-disruption-on-white-collar-metros as separate currents; the new sharpening is that AI-linked EMPLOYER relocation accelerates tax-base erosion — AI industry geography drives university geography through the employer/funding channel.

2. The "AI as accelerant of existing migration, not independent cause" causal framing with the specific mechanism: compute infrastructure follows cheap land/power/taxes south; university prestige follows industry funding.

3. Specific numbers: 20–35% SEC flagship enrollment growth 2025–2045; 25–35% college-age fall in NE/Rust Belt; 3–5x per-capita closure rate; one in four four-year institutions in that belt by 2045.

4. The endgame: regional brand monopoly — Southern flagships as default national universities; Northeast retains only elite tier; displaced students migrate south or exit degree-seeking.

a) WICHE 10–20% college-age decline (sourced, prior) vs. session's 25–35%. The 25–35% is unsourced session estimate. Prior consolidation note (Note 3 summary) explicitly lists 10–20% as the WICHE number. In Note 8's reconciliation the author decided: "I'll keep 10–20% (WICHE) and tag the larger cumulative figures as Unverified if used. Safer: use 10–20% WICHE."

I should follow that discipline: use the sourced 10–20% WICHE figure, and if I want the larger cumulative figure, tag it Unverified. Actually, I could phrase it as "college-age population falls 10–20% by the late 2030s (WICHE), with cumulative decline through 2045