I want to reiterate that I'm only against
mandatory arbitration agreements. I'm not opposed to fairly negotiated arbitration agreements or a
post-dispute decision to use arbitration. One thing I can do on the website is to make it more clear the scope of the problem.
> I'll restrict my arguments to commercial and consumer mandatory arbitration agreements. Those are the ones I'm familiar with.
This is definitely the area that has the strongest arguments in favour of arbitration (even mandatory pre-dispute). I'll have more to say later
> Every major arbitration venue has strict rules on who is qualified to arbitrate.
For example AAA and JAMS (which are the better known and more respected ones) do have standards. However, in virtually every study conducted they demonstrate both theoretical (arbiters saying that they are aware of who pays them) and actual bias (looking at the outcomes themselves).
> And the cost, speed and access is definitely better than litigation. (Delaying a case and bleeding out your opponent are separate. The capacity for both is much higher in litigation.)
This is only partly true. For example a company can spend similar resources to avoid arbitration in the first place. Also don't forget the mandatory "pre-dispute" negotiations that require you to appear or deep limitations on what's allowed (see https://gopro.com/en/us/legal/terms#dispute for one of the more egregious ones).
If the companies act in good faith and come to the table, and the disputed amounts are low enough that a lawyer doesn't need to be involved, this is probably cheaper and better.
> That said, I'm massively in favour of arbitration reform. For consumer--and commercial agreements where there is a material disparity in size between the parties--there should be a standard set of terms, written in plain English, which drafters may use. Step outside those, and the standard for judicial review (and state regulation) is lowered. If a sufficient number of arbitration claims are similar, they should be allowed to petition to form a class.
I don't yet have a fully developed thought for what the world should look like but here are some of my "early thoughts" to post on HN (as opposed to the website which I try and keep well-sourced).
1) Southland Corp. v. Keating should be overturned entirely. It, frankly, is unjustifiable based on the plain meaning of the law, the rules of statutory interpretation, and the historical context of the time. Fixing this would also return power to the states to decide their own contract law.
2) I'd love to see some system similar to Canada's civil resolution tribunal. This is effectively a "small claims" court that allows consumers to meaningfully get resolution without the need for lawyers or detailed procedure.
3) Consumer arbitration should be a revocable until arbitration begins. Recall that the CFPB found that when signing these agreements only 7% of those that understood that they permanently gave up their right to sue. I'm not opposed to arbitration as a litigation strategy (which may be positive for both sides) but it must be a meaningful choice.
4) Perhaps give better access to "limited jurisdiction" courts to consumers. This allows for a proper lawsuit but without the complexity of "general jurisdiction" courts. Most of the delaying tactics are more restricted in such a court.
5) Recall that arbiters are not required to (a) know the law, (b) apply the law correctly, or even (c) explain themselves. This is broken. I don't know of a good solution that would retain the speed and "finality" of arbitration but allow for meaningful oversight.
Class action waivers are insane and should be disallowed on the basis of public policy. I can't find any argument against disallowing them.
I'd also love to see terms of service as a whole be standardised. Similar to your suggestion about arb agreements - but for terms as a whole. You should be required to use some kind of "standard paper" or be subject to severe judicial oversight that predisposes against the drafter.
FWIW I did actually have the "cost, speed and access" arguments listed on the "problems" page -- though I could probably spend some time to expand them to make them stronger.