18 August 2011

A good comment on TEoA

Chris Perry wrote me this interesting email:

I enjoyed your book -- many good insights and entertaining/revealing stories.

As background, I worked for many years for the World Bank as an economist specializing in water/irrigation, was subsequently head of research at the International water Management Institute, and am now a part time consultant and co-editor in chief of Agricultural Water Management. I know and have worked with several of your colleagues in and around Wageningen.

You write about topics that I have also written about in various papers, and while I had many comments (supportive and questioning) as I read the book, I would like to concentrate on three points, which I try to make briefly.

First, you have avoided the primary error that many writers, and especially economists, make when analysing water issues: you distinguish between consumptive use (where the water applied to a particular purpose is evaporated or transpired by a plant into water vapour -- thus rendering it unavailable for further local use), and non-consumptive uses, where the water mainly passes through the using system and returns to the environment for re-use. I am not sure though, that you fully appreciate the extent to which non-consumptive use is prevalent in non-agricultural water. At one point you mention that the water going to a household is really consumed only when it is drunk -- a use that tends to return to the system pretty much intact, in physical terms -- and later, when you write about excessive housing development in water-short area of the US, you could significantly strengthen that case by adding that suburban houses (with large gardens) consume a far higher proportion of water abstracted than do urban houses.

Second, and here I really worry about how readers from development agencies might interpret your book: the vast majority of the examples you give where novel market-based solutions could lead to better/more productive allocation of scarce water are where water rights are established and enforced. This, of course is an absolutely critical precondition for markets to work. Related to this, I note in the last chapter you make passing reference to the need (while all the "stakeholders" are debating optimum allocations) for a basin-level oversight agency to make sure that hydrological reality is embodied in their decision processes. The absence of such authority, especially in cases of over-abstracted groundwater, is in my view a far greater barrier to sound water allocation and use than the absence of markets to allocate the stuff. In a paper[1] written some 14 years ago at the height of the Dublin principles/water as an economic good excitement, we concluded:
our position can be stated more positively, and to do this we define a necessary and sequential set of preconditions for the beneficial introduction of market forces into the allocation of water, namely that:
  • the entitlements of all users under all levels of resource availability are de- fined and include specified assignments to social and environmental uses
  • infrastructure is in place to deliver the defined entitlements
  • measurement standards are acceptable to the delivering agency and users
  • effective recourse is available to those who do not receive their entitlements
  • reallocations of water can be measured and delivered, and third-party impacts (in quality, quantity, time, and place) can be identified
  • effective recourse is available to third parties affected by changes in use
  • users must be legally obligated to pay defined user fees through effective legal and policy procedures
  • large-scale transfers of water with and between sectors must be subject to approval and relevant charges by regulatory agencies
With these sequentially interdependent preconditions in place, we believe that the privatization of water allocation would have significant benefits; in their absence impacts are uncertain. We also believe the implications of running experiments with peoples’ livelihoods, especially where water is involved, to be unacceptable.

Our contention is that development and the efficient use of water will be better served by the widespread, indeed universal, introduction of the necessary underpinnings and prerequisites to good water management (assigned water rights, delivery of a defined service). With these extraordinarily difficult steps in place, further pursuit of market forces in the allocation of water will be useful.
I think this list remains valid, is absent in most countries where water is scarce and contested, and is rather "assumed" in your analysis.

Third, even if these elements are in place, water remains a difficult commodity to marketise.

Pricing household use leads to adoption of low-flow showers and WCs with low flush volumes --but neither of these uses is consumptive so if we are trying to release water for the environment, wetlands or agriculture such interventions are of limited significance. (I resume you have read Gleick's ideas in relation to California, and the rebuttals by Charles Burt of CalPoly?)

Agriculture is different. As in the case of domestic supplies, what we can readily measure is water use, that is the water that is abstracted from a river or aquifer and delivered to the irrigated area. What we want to influence is water consumed, which is some fraction of water use. In the case of surface irrigation, the proportion is often quite small -- less than 50% -- which leads to the large and misleading literature describing irrigation as wasteful and inefficient. In reality, much of the non-consumed fraction returns to the environment for re-use at another time or place. But the farmer faced with rising prices for water or the option to sell his diversion entitlement to another user will look at technical options to maintain consumptive use while reducing diversions. This is exactly what hi-tech irrigation technologies do. The fraction of diverted water that is consumed may then rise to 80-90%. (Typically, with hi-tech irrigation, the plant's water needs are met more precisely, so that productive transpiration increases, along with yield). The farmer can divert less while maintaining constant irrigated area, or alternatively he can expand irrigated area, and hence expand total consumption. Such cases are documented and present us with difficult problems if prices and markets are to be adopted as approaches to dealing with scarcity.

The recent dispute between Wyoming and Montana the recent Supreme Court decision regarding a Montana/Wyoming dispute is an example. As I understand it Montana asserted that Wyoming, by converting to sprinkler irrigation, was reducing flows to Montana because even though it had enlarged neither diversion rates, volumes nor irrigated acreage, by converting to sprinkler it had increased consumptive use and hence injured Montana. The Supreme Court found for Wyoming. (As someone else has observed, "you can take a judge to water, but you can't make him think")

Similarly, in the Snake River Basin in Idaho, it was discovered that the long standing policy of benign neglect had led to many cases where farmers had converted to sprinklers and applied the apparent "savings" to additional acres. Generally, viewed from the "consumptive use" perspective, what depletes the resource is increased acreage; instantaneous diversion rate on which most rights are based is virtually irrelevant.

These two examples (drawn from recent contributions to the Winrock Water website by Bryce Contor) point to the fact that pricing water deliveries encourages farmers to "improve" technology by switching to drip or sprinkler, "saving" diversions. These technologies result in a much higher proportion of diversions being consumed -- so overall consumption increases and scarcity is exacerbated. It is hard enough to measure deliveries in the typical conditions of many water scarce countries -- small, parceled farms served by crude infrastructure operated by poorly functioning, low-paid staff. Measuring consumption is far, far more difficult.

Until we are close to resolving those issues, I fear that many of the desirable interventions you map out in your book are diversions from the exceptionally difficult task of constraining water consumption to something approximating the renewable supply.

Australia is in the forefront of efforts to move towards markets. Don Blackmore was the CEO of the Murray-Darling basin during much of the period when that process was under implementation; I attach a letter[2] he and I wrote to the Economist following their very market-supportive survey of water issues in 2003. Neither he nor I hold the positions we did then.

The Economist's 2010 effort was much better.

To which I responded:
  • I agree with everything you say -- my original postdoc @ UCBerkeley was targeted at reporting water trades. I soon found a deeper problem -- markets were not going to work (at all or even well) without proper rights/tracking of g/w.

  • I've disagreed with Gleick's pronouncements on g/w and "efficient irrigation" for awhile; see this and this.

  • I've read and internalized many stories like the ones you cite on WY/MT and "efficient" irrigation. I tried to include those ideas in the book, but perhaps not with the emphasis that you give them, e.g., I led a workshop on all-in-auctions at the State Dept last month. Necessary conditions: Rule of law and quantified water rights.

[1] Perry, C. J.; Rock, M.; Seckler, D. 1997. Water as an economic good: a solution, or a problem? Colombo, Sri Lanka: International Irrigation Management Institute (IIMI). v, 16p. (IWMI Research Report 014 / IIMI Research Report 014)

The Editor,
The Economist
July 23 2003


Can we first congratulate you on attempting to bring together information on such a complex issue.

However, your special survey on water (“Priceless”, July 19, 2003) embodies in its title a prejudice that experience from the real world rarely justifies. You refer specifically to the experience of the Murray-Darling (M-D) basin.

In the M-D, water use is constrained to equal the sustainable supply through a complex system of water rights, defined in terms of volumes and security of supply. In this drought year - the worst for more than a century - many users are receiving less than 16% of their “normal” entitlement, and that restriction is enforced entirely through the water rights system - not through pricing mechanisms.

Formally codifying these property rights - in systems that were already well managed and orderly; where customers were educated and accustomed to following rules; and allocation rules were already broadly in place and enforced - took a number of decades.

Once this process was complete, it was possible to introduce a system of trading in these codified property rights, allowing managers the flexibility to better manage their enterprises (in some areas last year as much as 80% of water delivered was traded). The water rights system also provides the basis for improved environmental management.

The parallel system of charging for water services in the M-D is quite separate from the sale and purchases of water rights, and exists to ensure that the income of water supply agencies is adequate to cover ongoing maintenance and projected major capital replacements.

Three lessons may be drawn from this successful achievement of sustainable financial management and sustainable resource use:

First, the primary means of balancing supply and demand for water resources is definition of water rights consistent with available supply. This is the approach followed in Australia, Israel, the US, and elsewhere.

Second, defining water rights is contentious and difficult at the best of times. Where water is already over-allocated so that "tail enders" often get no water, or fresh aquifers are consistently overdrawn to meet current demand, defining and enforcing sustainable water rights is an enormous political and social challenge. This is the case in many water-short developing countries.

Third, the primary role of water pricing in irrigation is not to balance supply and demand, but rather to achieve sustainable financing.

Implying, as the Economist article does, that pricing water has a central role in achieving the required resource balance is to grossly mislead policymakers facing the challenge of reducing water consumption to a level consistent with long term availability and proper environmental management. The solution inevitably requires stable and well specified access rights to water, institutions with the capacity to manage the water access regime, and appropriate water pricing to ensure the long term operation of the infrastructure.

Don Blackmore
Chief Executive
Murray-Darling Basin Commission

Chris Perry
Professor, Economics of Irrigation
Cranfield University
Cranfield, UK


  1. This is definitely one of the best posts ever on this blog. Thanks for posting.

    It reminds me that I need to better take into account points 1 and 3 in my research papers.

  2. David,

    Thanks for posting. Intriguing read. Do you know what Economist articles Chris is referring to in his letter?


  3. @Mark:

    2003: http://www.economist.com/node/1906846

    2010: http://www.economist.com/node/16136302

  4. Regarding this statement: "Pricing household use leads to adoption of low-flow showers and WCs with low flush volumes --but neither of these uses is consumptive so if we are trying to release water for the environment, wetlands or agriculture such interventions are of limited significance."

    Those are not consumptive uses and they will not affect water available for the environment within a basin. But when we talk about inter-basin transfers--as is the case in much of California--a reduction in out-of-basin demand (even in non-consumptive uses), frees up more water for the exporting basin.

  5. Can we get an expanded discussion of consumptive versus non-consumptive use? For instance, water that evaporates but that then becomes rain higher in the water shed might actually be non-consumptive while water that escapes from a water system into the ground only to trickle through the ground very slowly seems to be a consumptive use. It is not clear to me, how to include either of these cases in an economic analysis.

  6. @Eric -- the difference depends on scale and time, i.e., Farmer A's runoff is "non-consumptive" if Farmer B can use it immediately bur consumptive (wrt Farmer B) if it reappears later @ Farmer C's land. These are economic -- not hydrological -- definitions...


Spammers, don't bother. I delete spam.