We had an interesting recent economic policy discussion here at QUT about the topical issue of urban water management, chaired by Clevo Wilson, a senior lecturer in environmental economics. The full presentations can be downloaded here. The essence of the debate was whether it would really make sense to try to use price signals to reduce water usage in the big cities like Brisbane, Sydney and Melbourne, a popular topic in econblog land with Andrew Leigh being a vociforous supporter of more aggressive urban water pricing (see, for instance, here).
levo Wilson introduced the background issues involved. The generic issue is that in the region of the major cities, rainfalls have decreased in the last 30 years whilst populations and the water usage per person have massively increased. Whilst urban water users are responsible for no more than about 10% of total human water consumption (the rest goes into agriculture and industry), the urban water system is in many cases geographically separated from the other 2 main users of water in Australia and only at prohibitive costs could they be integrated by long-distance pipelines and the like. There is only a bit of farming that competes with cities for the same water. Clevo explained the main issues in urban water management to be the existence of a natural monopoly in terms of the distribution grid; the variability of demand and supply making marginal costs fluctuate widely; and the existence of other users of the same water (ecosystems) who were not well-funded and would not be able to outbid a city for water. Clevo sketched a water system that was in crisis with strong environmental consequences. There are two basic possibilities for limiting urban water demand: using price signals or appealing to social norms and restrictions. At present, we mainly do the latter.
edzo Mujcic took the classic economic position that we should rely on price signals. Redzo noted that water-tanks were economically inefficient because the implicit value of water out of a water-tank is probably in the region of 2$ per Ml (=1000 liter) , about twice the actual price and above the price of generating water from desalination plants (which cost about 1.70$ per Ml). Redzo found that the literature suggested the demand elasticity to be about -0.4 implying that any serious reduction in water demand would need pretty draconic increases in water prices (you’d need to at least double the total prices of water to halve consumption). Redzo finally noted that a private water supplier would probably over-charge for water because of the natural monopoly, in which case the articifical separation of the water market for urban and proximate rural areas would not be sustainable. Nevertheless, Redzo’s recommendation was to enforce marginal cost pricing on all users of water simultaneously, including the bits of the rural coomunity competing with the city . Because the fluctuations in supply and the importance of geographical location, this would entail a great degree of price differentiation: users in elevated areas would pay more; water would sharply increase in price depending on the level in the dams; there’d be a difference in prices for guaranteed water over the price of water-if-available; areas with more leaky pipes or areas further away from the source would have to pay more; etc. Prices would not necessarily have to rise to meet marginal costs, but rather in order to ensure sufficient capacity for droughts.
emanja Antic on the other hand argued that we should simply give up on marginal cost pricing or any other rpicing scheme because it would be too difficult, be perceived as unfair, would be open to abuse (people illegally tapping into other people’s water), would be open to arbitrage unless we were willing to axe the part of rural Australia close to cities as a viable economic entity that did not recieved water at a lower price than cities. Nemanja’s argument was that an appeal to command and control was the only really viable option, and the most popular option by far in other dry areas of the world too.
I was particularly struck in this debate by the immense complexity of actually pricing urban water and by the quick realisation that the water price increase needed to elicit a sufficiently strong demand reaction would make desalination economically viable. Hence, flexible urban water pricing appears to be an urban myth in terms of being an easy and quick solution. Once we really start to get worried about urban water, we’re simply going to desalinate. As a humurous aside, I foretell that if we do desalinate (and Queensland is already building a plant at Tugun), then we’ll pretend the coal needed to power the electricity station that feeds the very energy-intensive desalination plants is not going to hurt the environment because of the tiny bit of carbon we can put back into the ground (carbon sequestration).
I’m not sure I follow the argument. If people reveal a preparedness to desalinate and the carbon externalities are internalised (as you say they should be) then isn’t a water market the best way to bring it about?
Hi Nick,
the issue with the water market solution is not really connected to the carbon effect of desalination. The problem is that many of the standard assumptions about markets are violated in the case of urban water: the drop of water does not cost the same thing for every neighbourhood in the city (the higher-up one, where the water needs to be pumped to, should pay more); there are increasing returns to scale in the distribution network making it infeasible to let a private party decide on prices (the monopolist could charge a fortune); the water property rights are hard to police (its relatively easy to steal your neighbour’s water or share in its benefit); optimal water prices should change regularly; the demand sensitivity to water prices is low, implying that one should charge far more than marginal cost to get a sufficient response; what one would want to do – auctions about future water streams conditional on weather – would have prohibitive transaction costs in a decentralised system.
Every time a water expert (which I am not) tells me about urban water, even more market deficiencies are brought to my attention. I just find it hard to see how you can make markets for water work in an urban setting. Charging a flat rate, desalinate, and using restrictions when supplies are low doesnt seem so strange anymore considering the problems with urban water markets. My limited understanding of the international evidence is that these difficulties have frustrated attempts elsewhere. Can you see a way to make urban water markets ‘work’?
“(its relatively easy to steal your neighbours water or share in its benefit)”
It is? How?
More generally, yes, the market will be a compromise – as is the market in all utilities (or all I can think of). But rationing? Well perhaps the elasticities make that politically impossible, since we want everyone to have their shower and wash up at reasonable cost and are politically constrained to do so anyhow, but it sure would be nice to allow people to trade their entitlements – don’t you think?
think about the transaction costs: do you really want Joe Average to have to trade their entitlements to 1000 liter of water on a Saturday evening in the event of rain in Oktober, whereby you lose 10% of that water if you sell it to your neighbour and 40% if it far away (loss during transport), with a 100 exception clauses?
Paul, I do not buy the argument about pumping water uphill being a significant cost. If you do the calclations then it turns out that a typical household 100m above sea level would use as much energy as it takes to run an average light bulb for 4 hours.
In any case, it is not essential to have the price perfectly adjusted to eh marginal cost. It isn;t now. The problem is that the price is just way too low.
What is wrong with progresively increasing the price of water across the board, after allowing x litres per person at a lower cost? The low cost water entitlements could be sold. No transaction costs that I can see.
You say that you would need to increase prices too much to get a response but you cannot know this. Elasticities cannot be extrapolated into unknown territory. And if you do, then that is the point at which the price signla makes desal kick in. Like Nick, I cannot really see the major problems.
What I was thinking about regarding trade is next door neighbours.
The rules one makes for rationing water are far more ridiculous than the inefficiencies of water markets. I mean one month you’re able to water between a couple of hours at night. If you’re not home then – tough. The next month you can’t use a hose, only a bucket. Then you’re only supposed to use grey water so people have longer showers with buckets at their feet – if they can be bothered even pretending to shower. I mean what is this – bush week?
Rationing just gets more and more arbitrary and ridiculous the further you take it.
You have got your units confused. 1000 Litres is a kilo-litre (kL). 1 million litres is a megalitre (ML).
Two small comments – firstly I can authoritatively assure you that State water bureaucrats in at least one jurisdiction are fully appraised of the arguments in favour of price signals, and they inform their Minister of the options quite regularly and thoroughly. It is not for lack of critical thought that these changes haven’t been made.
Secondly I wonder what the economic costs of desalination will be with a proper carbon tax price included? Somewhat more than water tanks, I would guess.
Paul, I don’t understand why you seem to think individuals will be trading water rights directly. What market works like this?
Take milk. the cost of milk distribution is not the same everywhere in the city – land and transport costs differ for starters. Similarly, there are increasing returns to scale in the distribution. But do you purchase milk straight from the farmer? Do you trade in milk futures in case there is a drought? If you are like me you buy it from the supermarket or milk bar and leave the pricing up to them. Water would be no different, except for the natural monopoly at the connection end, which, as Nick said, is no different to any other utility, and independent (preferably) government agencies setting pricing from different catchments according to their current water storage capacity, and future environmental requirements (in this sense they’d function somewhat akin to the Reserve Bank).
You are right that desalination would become a cheaper option, at least during droughts, but I doubt any city would have enough desalination capacity to cover all usage. Most likely desalination will provide a dampening effect on high prices (no bad thing).
I’m not so sure about the rural/city issue, in that I am not sure it is necessarily a problem, and because if rural industries are being favoured by legislative fiat, I don’t see how somebody could actually arbitrage water without the relevant authority noticing.
“Secondly I wonder what the economic costs of desalination will be with a proper carbon tax price included? Somewhat more than water tanks, I would guess.”
Not likely wilful. A while back I did some sums(JQs site as I recall) on our local council’s efforts to collect rainwater off the local civic centre and pump it to the downstairs toilets. In the range of $10-$11/kl as I recall and you’d expect they’d have more economies of scale than your average house. I pay about $1.10/kl from SA Water now. Those sorts of figures represent some real resources (remember the water is free), which is why we traditionally have large systems for those economies of scale. Rising carbon prices would be proportional you’d expect.
One of the big problems with tanks is your gutters are generally 2100-2700 mm off the ground while an internal tap spout is already 1000mm above ground. Hence the need for an automatic pressure pump, not to mention separate plumbing, since you’re not allowed to feed into the mains water for health reasons. Retrofitting is even more cost prohibitive. It’s all feelgood stuff for yuppies really.
As for rationing, new home builders don’t fit the meters to the site until handover as they often get nicked to be used for stealing water to fill pools and the like. Recently some Adelaide suburbs are having their existing meters stolen, although that could be for brass scrap now. Expect a rollout of plastic meters soon due to copper prices.
Chris,
I have to take issue with the assertion that pumping is not a significant cost.
If you contact your local water supplier, I am sure that they will provide figures on pumping, and they are significant. (Mind you, the sewage goes the other way perhaps).
Put another way, suppose you put a pressure pump on your rainwater tank and then supplied your house at 20m head, I reckon that even at one fifth of the 100m you were talking of, you would burn up appreciable amounts of power. Water authorities do in any event.
I also need to point out generally that most water suppliers charge costs for provision of headworks that covers at least the capital cost to supply difficult areas. This however gets built into the price of the house. So once built there is no further payoff for reducing water usage. It also means that some of the differential costs have already been taken into account since the water supplier did not pay for the increased capital cost of servicing difficult areas.
There may be some scope in the future for converting that to a per kilolitre charge for water with a commensurate reduction in the cost of the block to the buyers.
Ah, my memory, my lamentable memory.
I forgot to add re desalination:
Desalination is often scorned as a supply option because of the supposed high energy use associated with it.
However, it can actually be a lot lower energy footprint IF the option is used only as a peak and/or emergency supply. By that I mean if desalination is used as a supply only for the one in ten year drought and/or if other supplies fail, then the actual energy used is only in that one year in ten. Looked at from that point of view the actual energy cost and greenhouse gas production over time is limited. Looked at in comparison to other options (such as rainwater tanks whose manufacture is quite greenhouse intensive, or dams – enuff said) it might turn out to be relatively greenhouse friendly once other options such as demand management have been exhausted.
It’s a very strange argument that because we can’t charge perfect marginal cost for everyone that we shouldn’t try and approximate it for as many as possible. Sure, we can’t easily charge more for homes on hilltops, and sure an unregulated monoopoly would overcharge (but who’s proposing such a beast?). But these are just quibbles.
No-one is trying to get exact marginal cost on every litre – just to control total demand and stimulate total supply as needed by getting it right in the broad. That’s helped by the observation that in the sufficiently long run all costs are variable, so MC tends towards AC. So you do average cost pricing supplemented by temporary el nino levies or la nina rebates to address short term variations.
On political acceptability, water restrictions aren’t popular either. And a “tax” on water is likely to be highly progressive (the poor don’t have extensive gardens and multiple bathrooms) – something the pollies ought to be pointing out.
BTW Paul, if you do it by tradable quotas (not my choice) you will not require anyone to trade – just allow them to. The trade can then be as complex or simple as people like.
some pretty good arguments above, many of which I’d have to agree with.
Chris, you’re right about Ml;
observa, I agree with the water tank issue and never thought of the hegith issue associated with them;
derrida, the quotas you would have available to trade have to be contingent on a whole lot of factors (when and where it rains sinse you cant trade what doesnt exist, the type of use the water is meant for since with some uses a lot of the water can be recylced but not with other uses, the location of the person to be traded with, etc.); the nitty gritty of quota trading in water markets really is very tough and leads to transaction costs that would seem to preclude Joe Avergae from engaging in it. Big professinoal water users would probably make the investment and that’s for whom water markets are currently benig developed, but can you see your average urbanite trade water?
russ, you describe how a market for a normal good works which is fine, but water is not a normal good since there is a natural monopolist selling it to you who would heavily overcharge if unregulated, hence the ‘let the supplier decide’ market option is not available with water;
wilful, many governments have looked seriously at water pricing but few have implemented it why has your state governor, being so well advised not implemented it?);
marks, you are probably right that desalination plants would only be used during drought because it makes perfect business sense to first use the cheaper water falling from the sky. However, if water usage continues growing as it has, we’re be in a ‘drought’ permanently.
Moreover, if the desal plant is private or a PPP, how happy are the stakeholders going to be with sitting on a no-return investment every time it rains?
Well, “moreover” was utterly wrong, but anyhoo…
Typically, one would expect some kind of shared risk – ie a minimum annual purchase, which could be used to fill up reservoirs, for example. For example, power plants are often built on this basis, sharing the risk with either a power supplier or a major nearby plant.
Alternatively, a pension fund, private equity or very large utilities company (ie > $50b) might be happy to take the ‘lumpy’ returns in exchange for ‘lumpy’ returns if they anticipate monopoly pricing in drought years (which as a marginal supplier, seems reasonable to anticipate). They could store up water, for example, to increase their ability to exploit monopoly pricing opportunities.
Also, they would get significant upfront and ongoing tax deductions, which would help the overall project’s feasibility (they just have to be big enough to ‘use’ those deductions).
The real reason why it isn’t being done already is that there is too much risk – of the regulatory variety.
NB, you imply that you understand Paul to mean that the plant would only operate in ‘drought’ years – I understand him to mean that it would only produce (over the ten years) water for the one ‘drought’ year. Certainly that is how I would expect it to work.
The algorithms for working out how much to provide via desalination vs dams is quite complicated, and depends on what you are trying to achieve.
For example, if you want to minimise evaporation losses from reservoirs (which would allow that water to go down the rivers as environmental flows) then you might allow the reservoirs to go down faster so that there is a lesser surface area and hence evaporation. Then you would need to run a desal plant harder during droughts.
Another objective might be to minimise desal plant size and cost. In that case, you would run the plant at a constant capacity over the years. In this situation, it would need to be looking at the long term average deficit in say a ten year drought averaged over that time. A PPP would like this option of course.
Finally, if we get to the situation where demand management is worked to the full, and dams and desal plants are working hard, we might just be in a situation where we cannot support any greater population without making a decision to increase our carbon footprint to provide water.
If that’s the only decision, here’s my vote in favour, now.