PhD's in Math are very rare, and uneconomic. Aside from Wall Street and Langley, no one outside of SV Talent recruits are paying for someone who has spent their prime thinking years considering the viability of certain types of "up-my-sleeve" numbers - no one else has that capital for specialty, almost certainly fruitlessly, since infosec advantage isn't sum-net-zero. Any APT that will pay will have a slight advantage; that opportunity cannot be simply absconded.

That is why NSA skews the average with their hiring practice, let alone indirect contractor influences - although the pure math SME's are held tighter to the chest than even private contractors can boise.

> aren’t logically inconsistent.

Sure, if you remember PhD's in Ecliptic Curve Cryptography or Number Theory or applied but pure 'XYZ' field of promising arcane mathematics are extraordinarily rare, and skew towards a certain demographic as well. The motivated, undeterred, socially-inept few.

  And the former is similarly not evidence that they mostly hire people with edu emails.

As the excellence required increases, the numbers get low enough, you can hire ALL the talent. And have enough 'explanatory' budget left over for institutional-preserving things and normal bureaucratic neo-con noise.

You can hire more Math PhD's than anyone, and still mostly not hire math people.

There are very, very few Math PhD's that can, even theoretically, threaten the current risk portfolio of our nation. But if they even did exist, you would not want to signal them out by being the only one they hired.

All signals require noise. Work cannot be performed absent a temperature gradient.