19 Nov 2009

How much do subsidies cost, anyway?

JR asks:
What is the annual dollar value of water subsidies? I know Jimmy Carter was trying to end water subsidies and the guy who wrote Cadillac Desert [Marc Reisner] said that for every two dollars invested in water subsidies, one dollar of economic value was created. In any event, what is the annual tax payer cost of water subsidies?
I can't give an answer to this question, but here are a few thoughts:
  • The biggest subsidy to water users is the "opportunity cost" of water they use, i.e., the difference between what they pay (say, $20/af) and what that water is worth to someone else ($50--1,200/af). If you charge them an "unsubsidized, cost of delivery" price of $50/af, they are still subsidized!
  • Without a market for water, it's hard to know the value of what's being used. Even if there was a market, we would only know the value "on the margin," which does not capture the (inframarginal) benefits that accrue to users.*
  • Timmins (Econometrica, 2002) estimates that water managers in western utilities invest too much in infrastructure, such that they last dollar (of their customers' money) invested only produces $0.45 of benefits. This result is not directly comparable to Reisner, but it indicates the same problem is little bang for the buck.
  • The value of water subsidies/rights are often capitalized in the land that those rights are attached to. A farmer who pays $20/af for water may have land worth $5,000/acre while land at a neighboring farm without water is only worth $1,000/acre.
Can you folks say more? Do you have any calculations of the value or water subsidies, on any scale, at any place on the planet?**

Bottom Line: Subsidies distort prices and therefore distort decisions on actions and allocations of resources. If we want to maximize the value of water, we should end subsidies and distribute water through competitive markets.

* If a $1 can of Coca-Cola is worth $5 to you, your "surplus" is $4; if it costs $0.50 to make/distribute, then total social surplus is $4.50. That may be inframarginal; your 10th can of Coke may only be worth $1.01; there's value, but not that much. Social surplus can ALSO go to zero at the margin when producer costs are rising, unlike my example.

** This PDF (via SH) discusses subsidies on California's State Water Project, pointing out that (urban) MWD has paid 62% of its costs for 31% of its water, while (agricultural) Kern County contractors have paid 13% of costs for 42% of the water. Although MWD should have higher costs (pumping a longer distance, over the Tehacapis) and this is prepared by Public Citizen (a group that sometimes imagines nasties), the report does raise good questions (and give numbers!)


  1. The State Water project receives no money from the General Fund; the construction and operation of the facilities are paid for entirely by the users. For a variety of reasons, the allocation of cost within the user group is not done on an across-the board toll per acre-foot. Leaving aside the differing structural needs of the agricultural versus urban users (urban users need 100% reliability, treatment, etc.)I think the biggest driver for the present scheme was one of volume. At the time the project was being planned, MWD understood that if the agricultural users in Kern County were asked to pay so much that farming was uneconomic, they would not participate. By the same token, if MWD users were forced to cover the entire project cost, their water rates would skyrocket. Hence, the marginal income from a large block of water, even if provided at a lower price to the users, made sense. So there is really no subsidy either way. There was an implied subsidy in the granting of eminent domain powers to the SWP, and possibly in providing State land to the areas now used as reservoirs, but these are pretty small potatoes.
    The Federal system is more complex, and more of a typical handout scheme; but as I understand it the primary subsidy is forgiveness of interest costs on the repayment of project construction. I'm sure another poster is better informed on how that works. But I don't think it is a huge amount of money.
    You are right: the biggest cost is the opportunity cost.

  2. "Without a market for water, it's hard to know the value of what's being used. Even if there was a market, we would only know the value "on the margin," which does not capture the (inframarginal) benefits that accrue to users.*"

    Great comment, and one I'm grappling with in regards to the coming carbon market and offsets.

    Mister Kurtz, somebody had to pay between the amount that Kern demanded to participate, and the real value of the project. If that was the Met., then it was a subsidy. If that was the state, then it was a subsidy. Any time somebody offers to pay for somebody else's use of an item, it's a subsidy. Trying to hide it in the terms of the subsidy does not make the subsidy go away.

    David, I'm not in complete agreement with your bottom line. Sometimes competitive markets can create market distortions, too. For example, your second sentence in the quotation above.

  3. For another perspective on the dollar value of ag subsidies see http://www.ewg.org/farmsubsidies

  4. Josh, I'm not so sure about who's subsidizing whom. It really gets into the theory of marginal pricing, as when an airline will accept just about anything to fill those last few seats on the plane. If capacity were totally flexible, one could make the case that all prices for the same class of ticket (or acre foot of water) should be uniform. In the case of the SWP, the huge fixed costs of the system have to be paid back by the users. I could argue that the farmers in Kern County are subsidizing the MWD users, because the farmers, and the money they are paying, are the ones that made the system possible. The MWD users would be paying far more for their water if the agricultural users did not exist.

  5. @MK -- there was a subsidy in the sense of *some* organizations participated (had access to) the initial contracts. If the contracts were reSOLD every year (auctions!), the price paid would reflect the REAL value of water, which would include opportunity cost.

    Without doing that, you can still get efficiency by requiring (through all in auctions) SWP water contracts to be sold. In this case, the money would go to those who had the contracts.

    @Josh -- competitive markets can increase distortions, but they CAN be much more efficient than the so-called "solutions" that bureaucrats can impose. Just price in externalities and/or use common property law...

  6. good points DZ, but because of the very high cost of SWP water for ag users (I know, everyone believes this is cheap water; not so!), and the soil/climate profile in the better parts of the service area, growers planted trees. "Permanent" crops like trees crops have a more inflexible demand curve than annual crops like cotton, since you can't stop irrigating an orchard for a year and expect to have anything left the next year. The vagaries of an annual water pricing scheme (added to the natural vagaries of world markets, weather, insects) would have so depressed hypothetical returns as to discourage sensible long term investments in valuable crops and sophisticated irrigation systems. Other than that small quibble, I remain in agreement with you.
    The whiny comments by Public Citizen, and others who are resentful of the Sandridge sale, never cease to amaze me. Here is a land owner who paid the full cost of the water, including repayment of the bonded indebtedness, who could not make adequate returns from farming. He sells his water to another party who imagines he can put the water to a better use, and who assumes the risk on the transaction. In the twisted mind of the collectivist, there is a sin here somewhere.

  7. Mister Kurtz, just because a group of people threatened to not play unless they got subsidized does not make the subsidy go away.

    Perhaps, considering the water regime in the Valley, farmers shouldn't have planted trees. They were more lucrative when their participation was distorted by a subsidy for their involvement, and now, if we go to a more efficient system, their investments are lost. That is unfortunate, but sooner or later we've got to go to a more efficient system.

    If I were a farmer, I'd be spitting mad at the deal struck years ago that gave me the impression we'd have a good long-term thing in trees. Or, maybe I should have considered exactly where I was planting, and not done so, regardless.

    David, I completely agree, sometimes markets can be less efficient than government programs (I'm being mean, but this can be read from your statement). I'm having a problem with pricing in externalities right now in regards to the price on the margin vs. the actual cost of the effect of the externality, which I'll blog about soon.

  8. Also, David, your advice to "use common property law" is government.

  9. @MK -- we agree about water trades as a good way to increase value-in-use, but I'll stick with my subsidy line -- noting that it's a valuable property right...

    @Josh -- "use common property law is government." -- if you define gov't as courts, but courts (eg, arbitration) do not need to be gov't!

  10. I guess we're just getting into semantics here. The SWP was a commercial arrangement made between consenting adults, which received the blessing, but not the financial support, of the state. To my mind, when the government decides prices are too low for sellers of corn or solar panels, and takes money out of my pocket and gives it to them, *that's* a subsidy; or a more egregious one than whatever we are talking about with the SWP. The Federal water programs are different. They were stimulus projects, though pretty paltry ones by today's enlightened standards.

  11. @Josh -- it's a "public corporation" with co-operative management, taxation power and a significant presence on the bond market. You decide what it is...

  12. But for the existence of state and federal water projects, water from NorCal would simply flow out to the ocean without Californians being able to use it for agricultural and consumptive use. So the water projects provide a benefit to society. But who pays for the water projects? The water districts that are purportedly receiving subsidies -- still today these water districts are paying for the projects that were built years ago.

  13. I'll let the MWD define itself. From its website:

    "The Metropolitan Water District is a consortium of 26 cities and water districts..."

  14. @ Anon: at the risk of beating a dead horse, the users of the SWP have paid, and are paying, the entire cost of the construction and operation of the project. They are not receiving money from anybody. Joah and I are disagreeing (I think) about whether the negotiated scheme for allocation of expenses within the users of the SWP amounts to one party subsidizing another.
    Some of the users of the Federal project (the CVP) pay a respectable portion of the cost of that system, but not the entire cost, through a series of complicated arrangements I am not familiar with. That project, like the Interstate Highway System, was deemed a "public good", and our elected officials decided it was worth investing public funds. Just like all those other "shovel-ready" things that are doing so much to lower unemployment today.
    @ Josh: I believe the MWD and the Kern County Water Agency (and its subsidiary member districts) are legally equivalent, but I am not a lawyer. These water districts have power of taxation and condemnation, are funded by their members, and operate as public agencies under the rules of the Brown Act. Individual farmers do not negotiate with the SWP or MWD.

  15. A couple thoughts late the the discussion:

    First, water is hardly the only subsidized resource or product. Besides the obvious, oil, transportation comes to mind. Society has decided it's a good thing to subsidize roads, air travel and barge travel through public investment in infrastructure, lower financing costs, etc. Unlike the SWP, users of highways use them at no direct cost at all; the costs of construction and maintenance are spread over users and non-users alike. So why is everyone so concerned about water subsidies - other than the good point that they discourage conservation?

    Second, last year there was an auction of Chino Basin water planned, but it was canceled. Too bad - it would have been an interesting economic study. But it wouldn't have revealed the true cost of water because of regulatory impacts on costs, which I suppose are effectively opposite of subsidies. Developers are required by regulation to show a 20-year water supply for new development, and they would have driven up the cost for the opportunities the water would have provided them.

  16. @Laer -- water subsidies (implicit and explicit) benefit the few at a cost to the majority. Highway subsidies are not the same.

    The Chino auction failed due to monopoly interference with water delivery, which gets us to far different topics.


Read this first!

Make sure you copy your comment before submitting because sometimes the system will malfunction and you will lose your comment.

Spam will be deleted.

Comments on older posts must be approved (do not submit twice).

If you're having problems posting, email your comment to me