I gave this talk at PERC yesterday.
Here's the paper.*
Here are my slides [PDF]
Here's the one-hour MP3 of the talk [11 MB]
Two big things came out of this talk. The all-in-auction is appropriate for redistribution of water rights within wholesale water agencies, irrigation districts or water projects of a certain minimum size. The AiA may not be appropriate on smaller scales, where informal trades function pretty well.
Second, it's important to roll out AiA gradually, with consensus among water rights holders. That means that third party impacts are likely to be small, new distributions of water will not be disruptive, and price discovery will allow outsiders to see what's going on. (That's a problem for folks who want to hide the value of their water...)
I'll be updating the paper to include these points and other useful clarifications.
If you have more questions on all-in-auctions, then email me.
* Abstract: Reallocation of water from those who own the rights to those who have higher values can be win-win-win---helping the seller, buyer and society realize the value of an endowed asset, a productive input and a social good, respectively. Reallocation by bureaucratic methods is inefficient because of the difficulty in determining who "really" deserves the water. Reallocation by market methods overcomes this information problem but may suffer from illiquidity of under-participation and/or under-trading.
The All-in-Auction (AiA) mechanism minimizes these effects by selling all water rights in a bid-only auction that maximizes liquidity, efficiency and social welfare. The AiA does not take water rights because owners indifferent to selling can simply bid for their own water, at no net cost. AiAs are appropriate for use within organizations (e.g., irrigation districts or wholesale water agencies) that distribute water among members; members would decide when and how to use them. The AiA was tested in experimental sessions, where participation and endowment effects were observed.