OK, actually, lately I've needed to justify off-sides, intimate acquaintance with characteristics of (n): small (12 > n < 1,750) vs large (>5,000) dataset manipulations for descriptive purposes or predictive power, knowing poor μ (mooooo) ≤ 5% N is an axiomatic threshhold to modeling normal macroeconomic, politcially self-serving gears which determine  or characterize populations ≥ 500K. NB the gaps.

Personally, I've no reason to believe analysis of abonormal distributions produces affirmative "national" or totalitarian governing strategy. I'm not convinced that standardizing every business input and output is a requirement of economic and lawful security and business. (Here is an interesting interview in which participatns dispute the social-level function of commecial litigation. Ambiguous to say the least. The people need righteous attys rather than "economists", desperately.) Then again, I'm a democrat, willing and able to negotiate specious benefits and losses to effect economic stability, over all and over time.

So I realized, again: all those joker citations rely on *.gov compound estimates to formulate a story fit to print. And it is that methodoligical gate to the pasture that separates the ignorant (phil.) from, at least, the skeptical.

Diversity is the key to economic and political evolution.