The ultimate point of my comments here in general? Quotas and aid should be based off of socioeconomic status (SES), not race. Doing so will disproportionately help racial (and ethnic) minorities because they are disproportionately of low SES.
Does they refer to the ancestors, in which case they often were, or does it refer to the present day people, which also describes slavery in the US.
>they probably couldn't so much as identify the group that enslaved them
Could the present day individuals of more recent slavery given an accurate identification of the group that enslaved their recent ancestors? One which wasn't stereotyping, and which neither excluded slave owners or excluded non-slave owners?
>nor were they being denied the right to be represented until a few generations ago.
They probably were. Voting was highly restricted in the past.
Things like affirmative action do not exist as revenge or compensation for some past act, they are there to undo some of the still present damage from those acts.
In which case income/wealth/SES are the better metrics, as they ensure those who are damaged get help instead of relying on some other measure as a proxy for damage.
In America, black people suffered organized, systemic oppression for generations which prevented them from accumulating wealth. That fact is not well capture by just looking at someone's income, or even all of their assets. That's also a question of all accumulated assets of all of their family and close friends.
A recent Pro Publica piece covers this very well: "The Color of Debt: How Collection Suits Squeeze Black Neighborhoods", https://www.propublica.org/article/debt-collection-lawsuits-.... I originally heard it on This American Life: http://www.thisamericanlife.org/radio-archives/episode/573/s...
I also think Ta-Nehisi Coates' famous piece covers the history of systemic financial oppression well, "The Case for Reparations", http://www.theatlantic.com/magazine/archive/2014/06/the-case...
Why do you focus your attention entirely on the latter category and also include African/Carribean immigrants into the beneficiary group? Do asian children who are still suffering from wrongs done to their ancestors not deserve help?
This is why SES is the better standard as it helps those who still need help, regardless if it was due to past systematic discrimination or if some more recent event created the economic harm. (It isn't perfect as it only aids based on economic harm, but it is better than other measures which attempt to measure economic harm with some other proxy.)
Quite a lot of discrimination has happened against Hispanics, which are largely white. To say nothing of other ethic groups of whites.
That Hispanic is even brought up when discussing race shows the absurdity of most of these ideas as Hispanic isn't even a race. (Even better is the concept that there aren't races in nature, but populations, of which we artificially attempt to group based on easily identified phenotypes, but which breaks down under scrutiny.)
That's true in a fairly meaningless formal sense (which really reflects the history of the evolution of the excuses for why groups were discriminated against as "outsiders" more than substantive differences in the nature of the groups), but not really in substance. The official (e.g., Census) categories pose "race" and "ethnicity" as orthogonal categories, as a legacy of the outdated idea that the former are real biological groupings and the latter cultural (when, in fact, both are cultural constructs, which is even recognized in the way "race" affiliations are actually now defined -- by self-identification, not biology, except in the case of Native American/Alaskan Native, where, for political reasons, somewhat different standards are used.)
OTOH, the ethnicity scale that is posed as orthogonal to the race scale has only two values "Hispanic" and "not Hispanic", and the way results are usually categorized from anything using the notionally two-axis race/category system is usually one combined axis by race for non-Hispanics, and then with Hispanics as its own category on that axis.
In any discussions in any context not controlled by formal (usually, Census bureau) definitions, "Hispanic" is just treated as another category no different from "Black" or "Asian" distinct from White in a flat, one-dimensional space of categories, not an orthogonal distinction.