[xevb3k]

I find quotas in bad taste, and somewhat lazy.

There is clearly an issue, but if we attempt to fix it with positive discrimination, we just create more issues.

A better way to approach the problem might be increased scrutiny (for discrimination) on boards/companies with significant gender bias, and increased penalties.

This would be harder to implement of course, but might go further to actually solving the issue.

[rayiner]

There is a fundamental unanswered question in the area of gender discrimination, which is whether we can remedy past positive discrimination simply by removing overt discrimination going forward. That presupposes two things:

1) That you can actually get rid of gender bias among corporate boards that are almost entirely men;

2) That the system isn't path dependent (i.e. removing the positive discrimination will result in an outcome that is the same as would have been the outcome if the positive discrimination had never occurred).

Real-world systems are generally path-dependent. If a storm knocks down a fence, the fence doesn't go back to normal after the storm passes. If you bend a paperclip, it will stay bent after you let go.

Taking affirmative action off the table is tantamount to saying that society simply should not fix problems that result from the path-dependent effects of prior discrimination. Now that's debatable, but I don't think either side of that debate is "lazy."

FWIW, I think the California law is probably unconstitutional. (The Constitution doesn't always say what you want it to say!)

[xevb3k]

If a fence is knocked down you fix the fence.

Quotas suggest an approach that works more like “we know some fences are damaged in various ways, one way many of the fences are damaged is that they have water damage. So we’re going to replace 25% of the poles in all fences, with poles resistant to water damage.”

There’s an argument that if you show that there’s been gender discrimination, you should throw out the entire board and put a new one in place using a process more resistant to all forms of discrimination.

But applying a fix without strong evidence that discrimination has taken place does seem lazy, because you’re attempting to apply a blanket fix. It does nothing to address the underlying issues, and tries to apply a fix in cases where no issue exists.

[tptacek]

Clearly, if you knock a fence down, you can't put the fence back up simply by saying "it is now unlawful to knock fences down". So: if not quotas, what repairs this fence?

[xevb3k]

If you find evidence of discrimination, you take action. What action exactly is up for debate. You could even go so far as replacing the board.

But not all, all male boards will be caused by discrimination, and adding a few females members will not solve this problem in many cases.

[tptacek]

What do you do about the generations of discrimination (of "fences being knocked down") without repercussions? You can outlaw and punish the behavior now, but look around and notice how many fences are still standing. How do you fix that? That's the path-dependence problem Rayiner is talking about.

[_bfhp]

To what extent may we apply this principle? Was the promise of forty acres and a mule to formerly enslaved black farmers in the U.S. too "far"? Was the actual result of some land redistribution and the Freedmen's Bureau providing supplies and opportunities to freed slaves too far? Maybe you believe this. If not, could we establish where the threshold is? That is, what's the volume of harm-over-time a group would drop below before being disqualified for any positive rights?

[djrogers]

How can you implement penalties for gender bias without introducing a threshold for an acceptable division of representation? And once that threshold is established, how exactly does it differ from a quota?

[eli]

I guess the argument would be that you look at practices instead of outcomes, but that seems impossibly hard to implement

[xevb3k]

Let’s say you expect boards in general to be 50% male/female.

In case where this is not the case, and where it has not been the case for a number of years you might launch an investigation.

It might be that there are reasonable reasons in this instance. For example, the company was cofounded by two women or men, it’s a small company, and the composition of the board has never changed etc.

It might be that the company has a board with frequently changing members, but always selects male members. In that case, it’s easier to argue that discrimination has taken place. At which point you might launch an investigation and apply penalties etc.

[gizmo686]

The proposal is that a poor ratio results in stricter scrutiny, not a penelty. They would still need to investigate and find other evidence of illegal bias.