that is physically, economically, and environmentally correct
John F. Raffensperger, Dept. of Management, University of Canterbury, Private Bag 4800, Christchurch 8140, New Zealand, john.raffensperger@canterbury.ac.nz
16 Dec 2008, edited further 9 Jan 2009, 14 Jan 2009.
This work has been done in collaboration with colleagues from the Water Markets Research Group. The text below, and any errors or omissions in it, are my own.
Here are the slides and notes from my presentation of 30 Jan 2009, to the e-democracy.org participants. Thanks to everyone for attending!
This essay outlines a proposal for a constrained water market within a catchment, operated by a central authority who has responsibility to protect the environment and to allocate water to large commercial users. We take the realistic and common sense point of view that shared water is both a sacred vital life element and a commodity input to commerce. The goal is to require commerce to respect the shared resource, while allowing commerce to meet its obligations at much lower cost.
The work builds on considerable existing science in hydrological optimisation and auction theory. This work itself has been scientifically vetted. We are both pro-business and pro-environment.
We do not propose any change in ownership of water. We propose to allow trading only in the existing administrative right to use water which society has already granted to commercial users, i.e., consent. Control of the resource remains with the elected unitary authorities as it is now. The proposal is only for the very large commercial users, and has nothing to do with water for sustenance. Small users, such as city residents, can completely ignore this new system, and would continue to get their water as they do now.
The market would be run as a weekly auction. Users would bid to sell (actually to hire out) from their existing consents, or to buy (a short term rental) in addition to their existing consents. Using (1) the bids, (2) all available relevant hydrology data and weather forecasts, and (3) a list of required environmental flows, the auction manager would solve a computer model (hydrological optimisation) which finds the best way to allocate water to all users simultaneously, while ensuring that all environmental flows were satisfied, now and into the future.
The environmental flows do not require a price or a bid; they are a constraint under which commerce must operate. This system would ensure that commercial water use never violates agreed-on environmental standards. All flows would be simulated before users were given permission to take water. This is different from any existing water market, where environmental standards are in place, but not enforced, or enforced only when violated in the extreme. Our system guarantees environmental flows.
The available commercial water would move to its highest value use. This contrasts with New Zealand's existing first-come-first-served gold rush process, which has no means to re-allocate water, and which incentivises hoarding under the "use it or lose it" principle. Users would have lower risk and higher profits, and we can prove this. There would be no losers, because if a user did not want to participate, they would still have their existing consent (probably adjusted for weather conditions) for free. Users would buy or sell only if they thought it profitable to do so, just as they do with fertiliser and equipment.
Transaction costs would be close to zero, much less than the current $NZ5,000 to make a consent application, saving users and government huge administrative costs. Unitary authorities would be relieved of considerable burdens from trying to guess what users need, e.g., their soil moisture levels. The software would give environmental managers more understanding of the catchment and better control of the resource, such as more precise knowledge of where and when the catchment was fully allocated, and far more flexibility to adjust environmental flows.
Storage, dams, and reservoirs are not needed for trading to occur. The proposal here will work for any combination of surface and ground water. Participants would be the large commercial users, whether farmers (consumptive users), hydrogenerators (nonconsumptive users), or water managers representing urban areas.
We have considered potential market problems, such as collusion and monopolisation. We cannot prove mathematically that these will not happen, but we have considerable evidence that users would find market manipulation to be quite difficult, due to the large number of participants, the auction's centralised control, and the dispersed nature of groundwater. A well owner has the most control of water nearest the well, and less control of water further away. Even if users somehow controlled the available water, the optimisation process still guarantees the required environmental flows.
Water markets. Selling a life necessity. Putting a price on the sacred springs. Commoditising a precious part of nature for the sake of business. Putting profit above cooperation. These concerns come to some people’s minds when they start to think about water markets.
Some people don’t want water markets simply because they don’t like markets. Why allow markets for water? After all, look at the mess markets have gotten us into – wars, famine, promiscuity, bad breath, and bankruptcy. But whatever business you’re in, imagine that business were not allowed to buy or sell anything except what the Great Leader thought you deserved, and you were paid only what the Great Leader thought you should get. Welcome to North Korea.
Some people don’t want water markets, because they believe that poor people will be injured. First, we have to agree on what we’re talking about, so we know what we’re objecting to. This proposal has little to do with charging urban users for tap water. This proposal is not about charging poor people for drinking water. Nor is this not about slamming disadvantaged people in developing countries. It is about managing the commercial use of water by business, especially the 75% of the world’s water which is used by farming, in the developed world. Of this water, much is ground water. This proposal is about allocating a precious public resource, choosing much more carefully between the environment and users who take that resource to make money.
Some people don’t want water markets because they feel such markets would privatise a public resource. I have bad news for you: that pubic resource has already been privatised. Here in Canterbury, for example, water has been over-allocated in many places. It’s all gone already. Privatisation is a “straw man” argument.
Some people don’t want water markets because they don’t want to pay for water. We’re talking mainly about farmers, who don’t want to pay for a resource which they feel has been free. First, under this proposal, users’ existing consents would be grandfathered in, but scaled sometimes to match the available resource. Second, what they don’t realize is that they are already paying for water. First, they have to pay to get the permit. Here in New Zealand, that transaction cost is about $5,000, and takes months. Second, users pay in lost income when they can’t get more water. Third, users pay for deeper wells when they and their neighbours draw down the aquifer. Fourth, they pay for incredibly expensive concrete and steel solutions, such as new dams and pipes, to store and move water (unless, of course, they can sucker government into paying for it; don’t get me started there). Farmers are already paying for water, and heaps. The cool thing about the proposed water market is that farmers will make more profits – a lot more profit. Farmers will benefit from this system. They could sell their unneeded water. The consent process could be done literally in minutes. Their water-related risk would be drastically reduced.
Some people don’t like water markets because “it would put a price on water.” Yes, it would. The flip side is allocating water some other way, say, by paying the right politician or hiring a good lawyer or having a brother-in-law on the water allocation board. The current system is broken! We are letting business use up a precious public resource for free.
Some people immediately think of huge expensive water works. Maybe we can float the Tasman Glacier to Beijing? Maybe we can build a pipeline from Westport, NZ, to Perth, Australia? No, those ideas are absurd. Some people are saying that we can't trade water unless we have dams and reservoirs, and that used to be true, but is no longer. This proposal can be implemented pretty cheaply – no concrete needed, except perhaps for a modest office somewhere.
Some people object to a market when they are really objecting to living under a constraint. People don’t like markets because they don’t like working to get what they want. But we can’t make rain. We have to learn to live with the available water, and to do that, we need to manage it far better. Living sustainably means living under constraints, constraints on our behaviour so there is more for our children and more for the environment. With global warming and population growth, these constraints have suddenly gone critical. We have to live under constraints.
A few well-informed people object to water markets because water markets don’t work very well, anywhere, world wide. In fact, in some places, water markets have allowed damage to the environment and injury to people who were not parties to the trade. To these well-informed people, I ask you to keep reading. I recognize that several key issues must be addressed before water markets can work. This proposal directly addresses all of those issues.
I put a long list of objections in the Appendix. Over the past several years, I have given presentations about water markets all over the world. Often, I asked people to write down their objections to water markets on a questionnaire, and I collected them. Then I answered some of them right then and there. So if you object to water markets for any reason, please bear with me. I’ve probably already heard an objection like it.
We need water markets for the same reasons we have markets in radios and raspberries. Markets are good at allocating stuff. You can go to the store and buy milk. That’s very handy, and beats having to milk your own cow, or use that nasty “creamer” powder in our coffee. You can go to work, and be involved with selling a product or service, and you get paid. That’s helpful in paying the bills. We need markets because we all benefit from trade.
Sure, the tendency of markets is to push against every possible constraint, to tempt people to steal, to incentivise using up every last drop. (Of course, that happens without markets, too.) Markets are not perfect. They have to be regulated. So we have to assume the rule of law.
The world has two main uses for water: business and everything else. First, people, especially farmers, use water to make money. Second, the environment uses water, and people use water for recreation, culture, household, and many other reasons. These two broad uses overlap, of course, but let’s split it out for now. So we need a water market which can distinguish between non-profit use and for-profit use, and which can move the for-profit water to its highest value.
The available pool of water is relatively fixed. We can’t turn Canterbury, New Zealand, into a rain forest. We aren’t going to build a pipe from Lake Michigan to Tucson, Arizona. Let’s be realistic. What we can do, and need to do, is find a way to allocate the available water so (1) water is kept aside for the non-profit and environmental uses, and (2) the commercial for-profit water goes to its highest value use. If we do not set aside water for the environment, we’re screwed, because the earth is going to die. If water goes for a low-value use when a high-value use needs it, the water is being wasted.
We need water markets for one more reason, and this may surprise some people, especially members of the Green Party. All of us need to see the true price of water, so that we know it’s valuable. New Zealand held a mass whinging when the government suggested we use low-flow shower heads. I was embarrassed for this country. We can’t have water both free and priceless. You know in your own heart that if it’s free, you’ll probably use more than you really need. Why fix the leaky tap if the water’s free? If you see a price, if you lose money from the leaky tap, you’ll fix it. And you’ll put in the low-flow shower head of your own accord. Most critically, we need business, especially agriculture, to see this price signal. I’m using households as the example, but this proposal is really for the big commercial operations.
 Agriculture uses most of the water. From http://www.climate.org/topics/images/world-water-use2005.png |
 And the problem is getting worse very quickly. From http://www.fao.org/docrep/u8480e/U8480E3l.jpg |
At a higher level, we need to price water to protect the commons. If business people are allowed to take as much as they wish, without a cap, the public resource will be damaged. By imposing a cap, business is constrained. Business people will scream. That’s okay – no one likes living under a constraint. But we need constraints. Business is damaging the commons, and is injuring non-commercial users and the environment. Society has to cap that use in order to protect the commons. It’s not clear that a water market, per se, is the right mechanism, but we’ll see that the proposal here will protect our water resources far better than anything currently in place. The environment will benefit from this system.
Interestingly, business’ damage to the commons also hurts other businesses. When one farmer is spraying water on a hot windy day, it turns up in the newspapers, and embarrasses all farmers. When one farmer draws down the aquifer, his neighbour has to spend more to draw water from a deeper well. When one farmer wastes water, his neighbour loses the chance to use it productively.
We can’t have water both free and priceless. We need water markets to (1) protect the commons from over-use, (2) to enjoy the gains from trade.
World-wide, water markets don’t work very well. The most extreme case is Chile. In the 1980’s, Augusto Pinochet’s military government took the advice of its University of Chicago-trained economists and created a pure free market system. Deregulation worked in many of Chile’s market sectors, but not in water. Here’s how it was done: “Okay, now you can trade water.” That was pretty much it. As part of the implementation, Chile made a long list of mistakes, including bad law enforcement, bad ownership records, and letting a few firms grab water in a way that hurt poor people. Apart from the ineptitude and corruption, the surprising result was that very little trade happened. Very few people bought or sold water! The lesson is that simply allowing free trade does not create a market. Why didn’t trade happen?
The answer is transaction costs. Making a deal was just too hard. Let’s see why.
To get more water, a business user, say, Ann, goes to government and asks for it. Government is likely to say “No,” because water is over-allocated just about everywhere.
So Ann could decide to buy water from someone else, thinking that the offset should cancel, and government should approve. To buy water, Ann must find someone else who is willing to sell. So she spams all her friends and family with email about this, puts an ad in the paper, and even puts up posters down at the local pub. Eventually, she finds Bob, her neighbour. Ann has to negotiate with Bob as to the price, write a contract, and work out what to do if Bob doesn’t do what he says he will do. This is a huge transaction cost.
Of course, Ann likes Bob well enough, but this is business. So Ann and Bob probably have to hire lawyers to make sure everything’s okay. At long last, months after Ann started this whole process, Ann and Bob go together to government to get approval for the trade. The transaction cost keeps getting bigger.
Government also faces a big transaction cost, to decide whether to approve the deal. To make this decision, the government should get a hydrologist to consider the relative effects. The hydrologist should use a proper hydrology model, which simulates how much the local river is lowered, depending on the distance between the river and the user’s well. These effects will also change over time.
The graph here shows how the river is affected over time, depending on the distance between the river and the well that is taking the water. Using hydrology simulation software, hydrologists can calculate these impact coefficients, Fi,j,t, which is the amount that the river is lowered at location j in period t, if well i takes 1 unit of water now.
Suppose Ann is close to the river, like the 5 km curve in the graph, while Bob is far away from the river, like the 13 km curve in the graph. The trade can’t be one-to-one, because Ann’s impact on the river is bigger and sooner than Bob’s impact, as measured by the government’s river flow meter at the control point.
So government says, “No.” But the government hydrologist observes that if everyone in the catchment took less water, carefully timed depending on how far each user is from the flow meter at the control point, the cumulative reduction would balance Ann’s additional take of water. It would take a computer model to work out all the simultaneous effects. But at this point, Ann will have had enough of the whole thing, as anyone would, and so Ann is likely to take the government to court. The transaction cost is now going out of control.
We haven’t even said anything about how much water is involved. The point is that making the deal work is time consuming and expensive – i.e., existing systems for water trading have big transaction costs.
The results? First, water is traded only where it is controlled, with a network of reservoirs and pipes. So farmers like reservoirs and pipes because this expensive infrastructure gives them some control over water. Infrastructure lowers the transaction cost of transferring water from one person to another. Second, when trades happen, they’re for big quantities, and for a full season or year. Imagine paying $5,000 just to get to the grocery store – you’d have to stock up for a whole year at least. High transaction costs reduce the frequency of trades.
If transaction costs were somehow magically lowered to, say, 50 cents, farmers would trade small quantities of water every day or even every hour. They would better off in two ways, the lowered transaction costs and the more frequent trades. Lowering transaction costs results in increasing returns to scale. That is, users would gain if the transaction cost were changed from $5,000 to $4,999. They would gain even more by moving from $2/transaction to $1/transaction, because transactions are more frequent.
A few places, especially in Australia, have worked pretty hard at lowering the transaction costs. The electronic bulletin board, as in the Water Trading Register shown below, helps buyers and sellers find each other. But they still have to get government approval, and the users are still on their own to work out a deal. The government should ensure that the environment is satisfied, but those issues are difficult to manage, so the environment usually takes it on the chin.
With reservoirs, some regions have set up simple auction systems, and the auctions lower transaction costs further. Buyers and sellers know where to find each other, bargaining is simplified, and the auction manager ensures that the deal works. But the trades must be one-to-one, in both time and quantity. So that type of auction won’t work in a place with complex hydrology like the Canterbury Plains or the Ogallala Aquifer.
The lesson here is that transaction costs gum up the whole thing. If government or an environmental group wants to get more water for the river, it’s literally going to take an act of law to get it done. So water allocation is going to get done crudely, even in the thinly-traded water markets in a few places around the world. Every decade or two, a committee will get together and decide something like, “Farming gets 30%, hydroelectric gets 30%, the recreation people get 2%, …” Water is shared, so it can’t be traded one-to-one except with a simple reservoir. It’s too hard to make the deal.
The environment is pretty much on its own, and farmers are then set head to head against all the other users, especially what they perceive as those annoying environmentalists. Farmers adjust their irrigation on daily or even hourly basis, but getting water for the environment in such a detailed way is out of the question. None of this can take into account the detailed scientific knowledge that is available.
People have traded stuff for millennia. Yet, with all the knowledge out there, we have still not worked out a good way to buy and sell something as simple as water. It seems that we need an entirely new idea. Whatever we do, it needs to have low transaction costs. It must coordinate each user’s taking of water with other users over time and space. It must move water to its highest value, because anything less is waste. Finally, the new allocation system must have explicit protections for the environment.
To find a new approach for allocating water, we will turn to other complicated goods, such as electricity, which require similar coordination. How are these good allocated? What is the latest science in resource allocation?
The new approach is the smart market. A smart market is quite different to any kind of market that most of us see every day, but some industries actively use them all the time, all over the world. A smart market is an auction that is cleared with the help of a computerised math model.
As far as I can tell, the phrase “smart market” was used first (in this way) in 1982 by Rassenti, Smith & Bulfin. (The “Smith” there is Vernon Smith, the 2002 Nobel Prize winner in economics.)
 |
 |
 |
| Stephen Rassenti | Vernon Smith | Robert Bulfin |
These fellows were interested in managing airplane take-offs and landings. They wanted to solve the problem of airport congestion (which many of us have experienced!). They wanted to let airplanes bid for take-off and landing slots. The problem is that take-offs and landings have to be coordinated. It’s no good asking a plane in the air how much they’re willing to bid to land, especially if they’re low on fuel!
Rassenti, Smith, and Bulfin wrote a math model which found the best set of take-offs and landings, ensuring that every take-off matched a landing for a flight that the airlines wanted, and kept flights within the capacity of each airport’s runways. Now, the U.S. government is implementing a system like this to manage the very busy airports in the U.S.
But who owns the slots? For New York’s JFK Airport, the airlines think they own the slots, since they’ve been using them. The New York Port Authority thinks it owns the slots, because it owns the land. The Federal Aviation Administration thinks it owns the slots, because they’re the ones that actually give permission to take off and land. Obviously, this is going to have to get resolved before the market will work, and lots of money is at stake. We just need to get beyond the initial point.
Another smart market example is the modern power markets. When a generator puts power into the national grid, it has no way to know which user is buying the power. The generator is simply charging up the whole grid. Many generators together need to be coordinated, to make sure that the supply of power meets demand, and to make sure that the wires don’t overload.
The solution was conceived by Schweppe et al in 1988, in a book called Spot Pricing of Electricity, and then actually implemented for the first time in New Zealand in 1995. (See William Hogan from Harvard, E. Grant Read & Brendan Ring from the University of Canterbury, NZ, 1996. Much of this work was pioneered by Cantabrians!). Like the airport landing problem, a math model does the coordination, this time ensuring that supply equals demand, and that the wires don’t overload.
Generators bid to produce power, “If you pay me $1, I’ll produce a little power. If you pay me $2, I’ll produce a lot of power.” Users (wholesale power distributors) bid to buy power, “If you charge me $1, I’ll buy a lot of power. If you charge me $2, I’ll buy only a little power.”
It’s critical to see that these bids cannot possibly be matched one-to-one, just as groundwater cannot be traded one-to-one. The computer has a simulation-optimisation model of the power network. The computer model matches the supply and demand bids, ensuring that the power flows correctly over all the wires. If one generator makes power, several users must to be ready to take it, and the lines can’t be exceeded. So it’s not a free market, but rather a closely regulated smart market, managed with the help of a computerised math model which has been programmed with the physics of the power network.
Based on math that has been around since about 1949 (when my hero George Dantzig invented the simplex algorithm for linear programming), the computer calculates how much power each generator should produce, and how much power each wholesaler should take, and at what price. And it doesn’t find just any old solution. It finds a solution that minimizes the total cost of producing the power.
Some of you may be rolling your eyes here, as New Zealand’s electricity market has been quite controversial. In the U.S., people immediately think of Enron and the California power crisis. Actually, New Zealand’s system is envied the world over, and is actively being copied. This past year, despite low reservoir levels, the system worked, because business backed off when the price got high. California’s problems were ones of bad initial market design, which they have since solved. And why are prices high? It’s because reservoirs are low, not because power is allocated by a market. High prices will get people to use less, which is exactly what is needed. It’s uncomfortable, but it’s the right thing. Of course, that doesn’t stop government meddling with it.
In fact, most of the developed world is now using smart markets for electricity. These systems run as often as every 5 minutes (as here in New Zealand), with millions of dollars trading hands all the time, in robust and highly reliable systems. Math models coordinating complex auctions.
A modern electricity market works by having a market manager, who acts as auctioneer, and who accepts bids from generators and wholesalers. The market manager solves a math model which coordinates generation and demand, ensuring that all system constraints (such as line capacities) are within acceptable limits. The market manager reads the best solution from the computer model, and then tells the generators and wholesalers what to do. The participants will accept this solution, because the math model ensures that each generator gets at least their bid, and each wholesaler has to pay at most their bid.
Many governments now sell radio spectrum by auction. These auctions are quite complicated. For example, with cell phone frequency, a cell phone company could not make a profit if it got the license for only one city. It could make a profit if it got the license for several big cities.
Bids take the form, “I’d pay you $2 billion for the licenses for Chicago and New York. I’d pay you $4 billion for Chicago, New York, and Los Angeles, because having the three big markets is much better than having only two markets. I wouldn’t pay for any one city alone.” The government has to make sure that each city license is sold only once, while maximising the government’s revenue. The computer model does just that.
In airport landing rights auction, ownership of the initial rights hasn’t yet been decided. After the initial allocation, airlines will be able to trade with each other. In the case of the power market, the flow of money will always be from user to generator. With radio spectrum, the flow of money is from users to government.
Notably, Murphy, Dinar, Howitt, Rassenti & Smith (same Smith as before) experimented with a simple (actually, a very simple) computerised market of the Southern California water network. This was for a game, an academic study in experimental economics, done with students in a computer lab.
Their experiment included about 10 players, and only for one period at a time. It did not consider complex hydrology. The computer had to coordinate only flows through a small network of pipes. Nevertheless, the authors showed that this simple market system worked well. Despite having only a few players, the market quickly found a competitive equilibrium. To my knowledge, this was the first time that a smart market for water was proposed. Murphy et al also ignored protections for the environment, though they addressed that issue modestly in a later paper.
Now smart markets are in use for a wide range of commodities, including natural gas in pipeline networks (as in Australia), University course registration (at my alma mater, University of Chicago), even for better catering at Chilean schools. Each of these markets has complex coordination requirements. Gas purchases and sales must be coordinated to get through the pipeline, without bursting the pipes. University course registration should let students get their favourite classes, taking into account classroom sizes. School catering requires coordinating transportation, while taking into account scale economies.
Most of us never see a smart market. We don’t need smart markets to trade ham or haircuts, because we can trade these things easily one-to-one with ordinary tools of trade, like standard law enforcement, bank cards, cash registers, and transportation. No one else cares whether I get one of these things (though perhaps some of my colleagues are wondering when I will cut my hair). Third parties are not affected.
We can’t trade water one-to-one, because it requires complex coordination based on the hydrology and the environmental requirements. If one user takes a lot of water, other users may need to take less, and timed appropriately, because the environment needs some water. The proposal here is to create smart markets for water, while allowing complex hydrology and protection for the environment.
Unlike electricity, airport runways, or reservoirs and pipes, complex hydrology is not controlled, especially ground water. Real catchments have complicated systems of rivers, streams, lakes, hydro generation, and aquifers. Fortunately, as mentioned earlier, hydrologists have means for studying and simulating these real systems. The simulation programs calculate “what if” there is a drought, or a set of wells take a certain amount of water. The programs can determine the effect on the water levels, in particular places of concern like stream flows, aquifer height, etc. The first requirement for our smart water market is a proper hydrology model, which simulates how much the local river is lowered, depending on the distance between the river and the user’s well.
Is this information good enough? The data has considerable uncertainty. Some uncertainty comes about because scientists do not have access to users’ water meter records. Since we don’t know how much water users are taking, it’s difficult to calculate the effects of those water takes. This man-made variability could be eliminated with proper metering.
Government is already making decisions, with or without data. The goal here is to use the available data more effectively. As we will see, the smart market system will tell us precisely where the data needs to be improved.
But even with perfect data, hydrologists have no way to say who should get water, nor when. Hydrologists generally cannot tell society what we should do, because the simulations have no information about what society wants, especially the value that business users have for water.
Before water is allocated to commercial interests for profit, society must decide how much water.
Sometimes hydrologists have access to information about required environmental standards, when government has specified those standards sufficiently. For example, hydrologists would be likely to raise an alarm when coastal salt water intrusion was about to damage the fresh water aquifer. In some places, governments have not yet established good environmental standards. We will assume that environmental standards have been set, and we will assume that these standards are set in a precise and quantitative way.
Business people often feel ambivalent about precise environmental standards. On the one hand, clear standards cannot be fudged. Business people can’t pretend they didn’t know the standards. On the other hand, clear standards improve business certainty, because businesses understand their requirements.
In any case, once society chooses the environmental standards, government can get on with the problem of allocating the remaining water to business. The real problem for business, however, is the shared nature of the resource. If you or I take water, but not both of us, the environment is fine. If we both take water, the environment is hurt. We’re looking at the transaction cost problem all over again. If the resource were parsnips, ordinary market processes can allocate the resource. But the resource is water, interconnected with everything.
Thus, even with perfect hydrology data and clear environmental standards, government cannot allocate water correctly. Government has no way to choose which user should get the water, beyond crude “first-come-first-served” rules. We need one final piece to the puzzle: users’ value for water.
Given the above information and users’ values for water, we have everything we need to calculate the best way to allocate water. Some water goes to the environment, as determined by the environmental standards. The rest is allocated to commercial users. Business people use water to make a profit, so the water should go to the highest bidder.
If we do not give the water to the highest bidder, then we face a problem where User 1 could make $100 with a cubic meter water, but User 2 has the cubic meter, and is making only $60. So the country is losing $40. This is a waste of water! The two users would be glad to make a trade, but there’s no way to do the deal – until the smart market.
To get users’ values for water, we’ll ask them for it! We’re concerned only with the big commercial users – perhaps only a few hundred major users in a catchment. These are the people that take the most water and have the greatest environmental impact. Small users can be pretty much ignored.
From the user’s point of view, it’s almost as easy as TradeMe. These users will have to register to be allowed to trade water. They will have to meter their wells.
It’s not quite as simple as TradeMe, because water has effects over time, and farmers need to be reasonably sure of getting water for the whole of a season. So users must give bids for each week remaining in the growing season. For example, in the first week’s auction of the growing season, each user will give a bid for every week in the growing season. These bids give the future value of water, telling the market manager how much water to save for users in the future. In each auction in each week, users can change the current bid and future bids, if their needs change. The current week is treated as firm. Future weeks are treated as forecasts.
Given hydrology data, the required environmental standards, and users’ bids for water, you could imagine that a computer program should be able to work out the best way to give water to users, while ensuring that the environmental standards are met. That is exactly what the smart market will do.
Even though the water is not controlled with reservoirs and pipes, the natural flows in the aquifers have some flexibility to be optimized. The smart market uses hydrological optimization, and operations research techniques of linear programming. These mathematical methods find the best way to allocate resources, subject to constraints.
Following solution of the optimisation, the market manager would tell the users how much water they got, and at what price. The mathematics of the optimisation guarantees that buyers never have to pay more than they bid and often will pay less. It also guarantees that sellers always get at least what they bid, and often will get more.
Finally, the mathematics guarantees that every environmental constraint is satisfied.
Users will have accounts that list how much quota for water they have in each week. After the auction clears, the market manager credits and debits their accounts for purchases and sales, in water and money.
Gone! Finding a trading partner is easy – it’s the market manager. Negotiating is easy – just put in your bid on the internet page. The market manager makes sure that the deal is arranged, like a modern exchange system. Users are not buying and selling pair wise, but through a pool which the market manager supervises.
In Australia, the most active markets are trading once per season. With this system, users could trade every day!
Now we could allow people to buy additional water for the environment above and beyond the required flows. Users would be automatically compensated. We have eliminated the old system of convening a committee once per decade, and the endless conflict between interest groups. We can still do that if we wish, but the process will be much easier.
Environment Cartyruben (Ecart) allocates water in the Cartyruben region. Local voters have democratically agreed that the Akaria River should have a minimum flow of 10, as measured at the town of Slynew, and a minimum flow of 11 as measured at Grinleead.
However, local Cartrybians are concerned that the region’s three farmers, Alf, Bob, and Cal, are taking too much water. So Ecart investigated, and found that the river is too low. Here is a diagram.
The amount that the river falls depends on distance from each control point to each farmer’s well. Ecart tested each farmer’s well and discovered that when each farmer takes 1 litre/sec, the flow at river falls, in litre/sec as follows:
Slynew Grinlead
Alf 0.10 0.05
Bob 0.04 0.16
Cal 0.22 0.09.
So how much should each farmer take? Let’s do some analysis, with A, B, and C being the quantity that each respective farmer takes.
Flow at Slynew = 12 – 0.10 A – 0.04 B – 0.22 C.
Flow at Grinleead = 13 – 0.05 A – 0.16 B – 0.09 C.
There’s no way to “blame” any farmer for taking too much! No one does “the most” damage. To be within the law:
Flow at Slynew = 12 – 0.10 A – 0.04 B – 0.22 C ≥ 10.
Flow at Grinleead = 13 – 0.05 A – 0.16 B – 0.09 C ≥ 11.
Simplifying, the legal requirement is:
Flow at Slynew: 0.10 A + 0.04 B + 0.22 C ≤ 2.
Flow at Grinleead: 0.05 A + 0.16 B + 0.09 C ≤ 2.
We still don’t know who should get the water. We know only the result of a particular decision. For example, if each farmer pumps 8 units, both required flows are violated. Should each farmer get an equal amount? None? What would be best for Alf? For the flow at Grinleead? For the economy? These are real decisions that that society has to make, but government has no means to make them.
This could be written like this:
Maximize A + B + C,
Flow at Slynew: 0.10 A + 0.04 B + 0.22 C ≤ 2.
Flow at Grinleead: 0.05 A + 0.16 B + 0.09 C ≤ 2.
The result is that A = 17.1, B = 7.1, and C = 0. The total water taken is 24.29 units. Note that the computer solution ensures that Slynew and Grinleead each get their required flows.
However, Cal doesn’t get any water. Is this fair? Of course not! Taking the maximum amount of water is not society’s real objective.
We can write this problem like this:
Maximize A + B + C,
Flow at Slynew: 0.10 A + 0.04 B + 0.22 C ≤ 2.
Flow at Grinleead: 0.05 A + 0.16 B + 0.09 C ≤ 2.
A = B = C.
Solution: A = B = C = 5.55. Again, the computer solution ensures that Slynew and Grinleead get their required flows.
The total amount of water taken is only 3*5.55 = 16.67 units, less than the previous method. Is this a good idea? Absolutely not! It would give the lifestyle block owner an equal quantity as the hydrogenerator! This is absurd.
This trivial example shows that the sustainable amount of water depends on where and when it is taken. It makes no sense to think of a percentage of the available water.
Suppose each user has consent to 5 units of water. Suppose Alf bids $5/unit, Bob bids $4/unit, and Cal bids $11/unit. We can write the problem like this:
Maximize 5 A + 4 B + 11 C,
Flow at Slynew: 0.10 A + 0.04 B + 0.22 C ≤ 2.
Flow at Grinleead: 0.05 A + 0.16 B + 0.09 C ≤ 2.
The best solution is this: A = 0, B = 8.2, C = 7.6.
Alf sells 5 units of water, and gains 5*$5 = $25.
Bob buys 3.2 units of water, and paid $12.8.
Cal buys 2.6 units of water and paid $13.
All three users go away happy. Note that each user paid according to the bid. We can make this even more useful to users by allowing them to bid in steps, such as a bid to buy and another bid to sell.
We can get the pricing better for users by using marginal cost pricing. The computer calculates the improvement in total profit if the flow at Slynew or Grinleead were changed a big. That value is the correct price to charge the users.
The water market should be operated by the unitary authority which manages the catchment.
The unitary authority is required by law to manage the environmental flows. It cannot be left to users, because users have no incentive to manage the environmental flows, and will be tempted to break them.
Furthermore, the market manager may to scale users’ rights to take water, to match the available resource. That is the job of the unitary authority.
In short, allowing another party to operate an unsupervised water market would be the unitary authority’s dereliction of duty.
Business people are going to be quite leery of this system at first. Under the current “use it or lose it” policy, a water sale could be interpreted as “you don’t need it, so we’re taking it, thank you very much”. The “use it or lose it” policy has to go, and government has to guarantee that users won’t lose their rights in that way.
Government will have to fund the first of these markets, partly because the systems must be managed by the unitary authorities.
Beyond that, users will enjoy automated resource consent process which is done in a few minutes. This will eliminate their transaction costs. They will enjoy the ability to sell unneeded water, and will have greater ability to obtain new water, thus reducing their risk in drought and increasing their certainty.
The unitary authority supervising a water market faces two types of risk. Both are easily managed with proper market rules.
First, users’ rights to take water must be scaled to match the available water. If users have more rights than there is water, the market system will require the auction manager to buy rights back from users. This would be a reverse auction, in which government buys off users to protect the commons. Such a system is highly unlikely to be sustainable from a business point of view. The solution of scaling rights is easily implemented.
Second, the environmental requirements may be specified incorrectly. Fortunately, the computer output shows the results at each environmental control point, and can express that result in dollars per meter of head. The market manager should study the location with the highest monetary value first. This tells the unitary authority where to put its scarce research dollars. If the control point has been set too leniently, then the market rules will have to specify how that control point is adjusted. The simplest course of action is to allow a scaling mechanism to adjust users’ rights to the new standard. That seems appropriate when the market is first being started. Later, when the market has been operating for some time, the environmental standard could be tightened via user compensation.
What could happen is that in drought, the price of water would go up. Then people would complain that the market is bad, when in fact the market is helping to alleviate the real problem.
Before users got permission to take water for this week, the environmental flows would be simulated using the best available hydrology data, for at least a full year into the future. That would ensure much more reliable flows for the environment, which has to be good!
Implementation requires a proper hydrology model, specification of the environmental requirements, and training of users. A secure web site and accounting system must be set up. This is all software and information technology, which is considerably cheaper than concrete and steel infrastructure.
A couple of years ago, a research proposal was put to the NZ Government to set up a water market. The estimated cost on that was about $2.5 million, about the same cost as what would be required just to cost-justify a new reservoir and dam.
These objections were collected over many months of presentations to users, citizens, scientists, and government officials.
(a) “Privatises” resource. It puts a price on water. (b) Put a $ value on water. View water more as “economic commodity” when it is something much more important. (c) Incentives/cost for other parts of society and the environment. What are the consequences of water trading? What are the responses?
To our knowledge, water is almost everywhere recognised in law as a shared public resource. Governments retain the power to control waters within their boundaries. Few countries allow privatisation of water, but rather they may allow privatisation of the administrative right to use water.
Trading would by definition put a price on water, and ideally this price would reflect society’s marginal value for a given source of water (i.e., from a given location), suitably weighted by constraints to ensure sustainability.
A simple example can explain the gains from trade. Suppose one farmer Graham’s tomato yield would improve by $50, if he had 10 more cubic meters of water, and suppose farmer Hamish’s peach yield would diminish by $30 if he gave up 10 cubic meters of water. If Graham bought 10 cubic meters from Hamish for $40, then Graham gives up $40 cash, but makes $50 on his tomatoes, for a net gain of $10. At the same time, Hamish gains $40 cash, but loses $30 from lower yield on his peaches, for a net gain of $10. Both have gained. The water has moved to a more productive use. Unless one trader is under duress, a trade will occur only if both will be better off. Trade gives society economic benefits. If government prohibits this trade, water will be misallocated, i.e., wasted.
A good can be “important,” but still be a commodity. Oil, electricity, medical care, and food are all important, and the efficient markets in developed countries enable people to have such things readily. The word “commodity” may come across as cold or impersonal. This is simply the abstract language of economics, even though we all have strong feelings about how these materials and services impact our lives.
In fact, the lack of pricing is causing many problems with water. The scientific literature on this topic is quite long, and goes back decades. [Ref United Nations.] Without pricing, a user who desperately needs water and can get it may be next door to a user who has more water than he can use. A proper market would allow these two users to communicate and come to an agreement to share the water, while fairly compensating the one who gives up water. The whole point of water trading is to give much better incentives to society. Where water is scarce, its price to be high to incentivise conservation.
Loss of power over who controls use/allocation of water from political/local government/landowner dominated to an economically/financially dominated.
Even now, people choose how to use water. Governments can allocate water at a high level, but it is still private individuals and corporations who choose how to use it, and private people are using the vast majority of fresh water (world wide) to earn income. So water use is already financially dominated.
Markets work best when ruled by law. If people could steal without penalty, then no one would be able to buy or sell, and society would be anarchy. It is true that people will always be tempted to steal, but that is a problem of law enforcement. A proper market for water would require careful laws and careful enforcement of those laws.
(a) Equitable access to water – monopolisation of water and how to provide for smaller water users. (b) The potential for monopoly ownership of water. This should be a public resource for public good. (c) Monopolies controlling water. Monopolies taking over resource. (d) Water monopoly – water taken up by the richest and control lost. (e) Buying and selling – potential for business to monopolise resource? (f) Control by powerful players (not necessarily the regulatory authority which is the water manager). (g) A wealthy operator could captivate the market.
(a) Potential to place water ownership in hands of questionable corporations (equitable access). (b) Different markets have different ability to pay e.g. industry/community vs. producer.
Personalised number plates (for example) are traded with inflated prices. Prices which do not reflect any added value. Profit is made solely on their speculative values. Wouldn’t want people to profit on the trade rather than the use of water.
(a) Water is a public good – trading will allow people to get a wind-fall by selling a permit that has been granted as a privilege by the community. (b) Start-up issues of equity/windfall.
(a) New Zealanders are not at the point of valuing water in ways that might be amenable to a trading scenario. (b) Trade speaks of traders, market share of strong profit motivation versus other social concerns. (c) Values between competing stakeholders.
(a) The cultural values held by parts of New Zealand society mean that water trading is unacceptable to them. (Water cannot be traded.) (b) Seeing water as a ‘commodity’ may marginalise other perspectives such as Iwi/Maori. (c) Cultural Aspects – taking water from the land. (d) A trading regime MUST take into account all of the community values of water i.e. environmental, recreational, Iwi, public ownership, social values.
(a) People or animals that may presently use water, i.e., Whio (blue duck) and are unable to pay for it, and may lose out. (b) If there is a “price” on water then there is likely to be increasing pressure for use that maximises the monetary return. Such use may not be compatible with environment and community goals (especially longer term goals).
Recognition of environmental, social and cultural values in a trading system.
(a) If the water resource is already over allocated and the person who isn’t using their full allocation trades their surplus, it therefore over-stretches the resource further. (b) Ownership of water and if a user is paying but not using their quota this excess water should not be sold as users in lower reaches could over use and cause salt water intrusion. (c) Current inefficiency in use results in more water in river which in over allocated waters or those with insufficient instream flows, is beneficial to instream interests. (d) How does it provide for instream values, especially given scarcity is usually a precondition? Scarcity almost invariably means instream resources have already been stressed.
Currently, even without a market, government does not always monitor water use, and the environment is often damaged. Indeed, aquifers are falling almost everywhere in the world, and salt water intrusion is a serious coastal problem in many countries. These problems have little to do with a market or a lack of market. People are simply taking too much water.
Water trading would tend to increase the total amount of water used, because users could trade their unused allocation. This would tend to “over-stretch” the resource, potentially causing problems such as salt water intrusion on a coast.
The solution in both cases is to have specific protections for the environment.
(a) How does it deal with uncertainty/information deficit? E.g. often don’t understand groundwater resources until they are stressed or instream values may not be investigated until they are threatened. (b) Will it negatively affect groundwater of everyone.
(a) Instream interests don’t consume water, yet can rarely afford to purchase, even if instream interest is significant. (b) Lack of public interest concerns - worth more and means more competition with instream/intrinsic values. (c) More pressure on available resources -- water ‘worth’ more. More chance of full use over longer periods -- difficulty in maintaining instream values.
Trading will increase the total amount abstracted as ‘sleeper’ permits get traded to others – increases risk of rationing and/or exacerbates low flows in drought.
(a) Trading will result in intensification of land use causing adverse effects on water quality. (b) Point Impacts – bottlenecks of resource constraint.
Encourages the idea of using everything that “flows to sea” i.e. is it a bad thing that resource is left over?
Ability to easily buy water, distribution problems. Sell your water then later cannot buy it back as there is simply none for sale in your locality. Sought after suburbs are often “controlled” so land prices are kept high and new entrants have to pay up.
(a) Potential to set neighbour against neighbour (depends on the trading regime – could be only a perceived risk). (b) Productive use to non-productive use. Can I afford the trade for my crops? (c) Continuity of supply, user bases crops on availability of supply then does not get it. Inefficient irrigation or use because of plentiful supply.
(a) If you are in an over allocated water area, then if you introduce a trading regime there is a high risk this will make the security of supply worse. (b) Over allocate supply. Water will become more scarce. Increase the chances of getting rationing. Reduce value of land. (c) What if upstream person buys and denies access downstream? (d) Trading wars. May create inefficiencies. Some land may end up without water. (e) Water belongs/related to the land. Trading takes it away from the land for anybody to use for any purpose.
(a) Small holdings that are uneconomic couldn’t afford to pay very much. (b) Urban communities can afford to pay more than rural groups. (c) Affordability to smaller users (large units’ ability to pay more). (d) Water shifting – to non productive sector i.e. forms to turn -- flow on effect on land productivity. (e) Water as a commodity linked to high use value also implicitly results in an economic framework of water allocation that may marginalise less commercial growers.
There is no point in trading as we have an excess of water.
Existing users happy with system.
Lack of confidence in current management systems – already over allocated.
Council should be the water sharer to maximise the best use of the water and not have individuals controlling surplus to their needs.
There is opportunity for people to trade now (albeit not that explicit). We people like to have their own security of supply.
Is there water scarcity – or can we improve efficiency.
Most efficient use needs to be reviewed.
Clarity regarding other ways to manage the problem of water scarcity.
Could use more water.
Is the water available?
Over allocation issues? Would it all be up for grabs?
Encouraging water to become a commodity ($) – problematic concept of “ownership”.
(a) Who owns the water? If you own the land, what water goes with it? (b) Confusion between ownership/rights to use. (c) Property Rights. At the present time property rights are unclear. Until property rights have been properly established in law, trading will not work effectively. (d) Who owns the water? Although there are now ‘right’ holders – tradability increases their rights – especially term. (e) Ownership of water is not clear. Maori may have a legitimate claim to water resources and are likely to object to others profiting from it. (f) Recognition of existing users’ rights. (g) Assumes private property rights of the water.
(a) Varying security of supply, varying values. (b) Who is guaranteeing the supply? The farmer, the seller, the Council?
Given that a water trading scheme were put in place, many concerns are about its management and how well it can be operated.
(a) Administrative system that needs to be implemented. (b) Technology challenges – much higher level of management.
(a) Changing states of knowledge of environment. How does trading take this into account and does this produce a barrier to change. (b) Water trading per se is not likely to be a good management tool unless it is linked to a useful set of knowledge (hydrology, climate, land-use) that needs to be dynamic and account for change.
Complicated – policy requirements / RMA process -- how to track on-ground.
Examples where operated – does it work?
Trading water between individuals within permit limits.
Concerns about overall control over water trading.
Important to hold “RESERVE” to safeguard for risks of drought etc.
Importance of “Low Flow”.
Restriction of staying within one aquifer.
(a) “Last year” – the original allocation. (b) What is to be used as the base to start from?
The length of term of any trade swap length.
Cost of maintaining.
Reversibility.
Response: Yes, the attempt to provide a market for whale hunting was a stupid idea that did increase the destruction of whales. It did this by creating an incentive to hunt them and was totally counterproductive. That is a great example of how not to protect the environment. The move toward strict regulation and quotas was excellent, and has done much to improve whale populations. Any market approach alone that does not carefully provide for and take into account critical environmental factors will also fail. There are examples of this being the case overseas.
Our system is more sophisticated and has greater safeguards than New Zealand’s transferable fishery quota scheme. Our system relies on quotas (consents), and automatically adjusts these downward if environmental flows are not satisfied, like the fishery quota system. However, our system also runs a complete ecological simulation at every auction (e.g. every day or every week) to ensure that the environmental flows are satisfied.
Water use, left unchecked and unvalued as it is today, results in environmental degradation. This is why our approach is based on environmental factors that have been determined by professional organisations such as ECAN. Our approach is environment first and market second. Because of this approach we have a method of implementing a solution to water management. Our approach is founded upon explicit protections for the environment
Response: Yes, the problem is highly non-linear and dynamic. To ensure our model will work, it is composed of linear components and objects which interact together to model the hydrology. This is done with MODFLOW, a standard extensible package for hydrological modelling. The advantage of this approach is that in the future as better components become available or new ones developed, these can be inserted into the model without having to rebuild it from the beginning. Some of the new elements that we anticipate adding to the model would be weather forecasting and the future planned water abstractions of users.
Response: This depends on the ability of the hydrologists to model the interactions, and may be more or less important depending on the user. We think surface and ground water interactions can and should be modelled.
Response: Yes, this is true, but we have to start somewhere, and quantity is easier to model scientifically than quality. A quantity-only system would still allow the existing regulatory mechanisms for use in quality. We are actively developing a separate system for quality.
As much as we can, we try to make our system take into account all the existing available information, including democratically-chosen environmental flows, the physics of the hydrology, users' existing consent levels, and people's willingness to pay for water. Just as Forever Fair eliminates transaction costs for commercial users, it also eliminates transaction costs for government and environmental groups. Managing the resource would be easier, more precise, and more timely.
Response: Yes, in fact it could be solved either way. With our capabilities with the supercomputer, we intend to join the researchers at UC that are modifying off the shelf packages so that they match the sophistication of our new approaches to solving problems. However, colleagues at Lincoln University and other recent research have convinced us that much simpler computation methods (the "eigenmodel" approach) would suffice. We have a refereed conference paper with a model of the Selwyn catchment using the eigenmodel method.
Response: We agree that any water allocation system must provide a way to manage the resource and we recognize the need for more information to do this successfully. This is why we start with a hydrological model to ensure that this management capability is present. The basic approach is classical hydrological optimisation, which is already a well developed science for related hydrological tasks such as contaminant remediation. The system will give everyone better information than the existing approach, because EVERY TAKE will be forward simulated. The output of prices shown by location across the catchments will be of huge help in determining where water is most valuable, and how much should be spent on infrastructure such as dams and races.) Council retains the right to set consents and to control all environmental flows. Indeed, our system depends on the council's authority in that regard, and our system would give much more precise information about when and where those consents can be set, which would greatly ease the work required for Council to do its job.
The 1992 International Conference on Water and the Environment in Dublin concluded that “Water has an economic value in all its competing uses and should be recognised as an economic good.”
From World Resources Institute, http://www.nutrientnet.org/nn.cfm?p=a202, 29 July 2005:
“Trading is not a substitute for a regulatory framework. It is a policy tool that can be used alongside regulation. In the United States, it is usually considered within the Total Maximum Daily Load (TMDL) process under the Clean Water Act.
“Trading is not a way of letting a market decide the environmental outcome. It is a way of using markets to achieve water quality goals most cost-effectively.
“Trading does not let polluters off the hook. It does allow sources facing high pollution reduction costs to purchase less costly reductions from other sources; but the reductions must be made and paid for.
“Trading does not exclude other policy approaches. Trading can be used in conjunction with other policy tools such as targeted subsidies, mandatory Best Management Practices, or public education campaigns.”