Our GMD is largely irrigation use - 97% plus. It then follows that if water use is to be reduced, working on a 10% reduction in the remaining 2% of usage is probably not the best approach. As such, we focus on the irrigation uses.
Traditional wisdom says "let's get more efficient with our irrigation use - we'll pump 20% less water." But, unfortunately this is not always going to be the case. Converting to higher efficieny irrigation can actually INCREASE water used by crops for production - even when less water is pumped and applied. Without a good understanding of what consumptive use was happening under the old irrigation systems, and specific constraints on irrigated acres and cropping patterns and irrigation management under the new systems, the conversions are not likely to save as much water as you calculate - if they'll save any water at all. They are, however, usually good for the producers' bottom lines.
Next usually comes: "Let's just regulate a 25% reduction in all water use." That'd work, but all water rights are different. 25% from the full water right is different than 25% from the partial right. And do you want to take this reduction from municipal, industrial, domestic and recreation rights as well? What about priority? Does the most senior right concede the same 25% as the most junior right? In Kansas where a water right is a property right to the use of the state's water, are there "takings" issues to be considered in a mandated reduction?
Usually someone gets around to the appropriation doctrine itself and calls for a 25% reduction based solely on priority. What happens if the most senior water rights to be protected are in the poorest part of the aquifer?
The market system usually gets air time, too. Water banking, water right auctions, transfers to higher economic uses, etc. But these can have significant impacts on the social fabric of the region because fundamental uses of water can change so much.
One thing is for sure - reducing water use by design is always about deciding if the water or the economy is more important, and then picking your poison. There is no way to change a deleterious trend without changing something about the system that caused the trend you want to change. This is why it's so hard to do.