29 Sep 2011

Time to dump increasing block rates?

In "Distributing Water's Bounty" [PDF], Ron Griffin and James Mjelde argue that increasing block rates (IBRs) are less fair to the poor than Uniform Rates (URs) [1]. Their argument goes like this:
  1. Start with a $10/household fixed service fee plus $4/unit volumetric charge equal to the delivery cost.
  2. Assume a poor household consuming 2 units pays $10 plus $4*2 ($18); a rich one consuming 6 units pays $10 plus $4 *6 ($34). Total revenue = $52 covers total cost of $32 + $20.
  3. Now create an IBR by reducing everyone's price for the first 4 units to $2/unit. The rich don't use more [2], but the poor household now uses 3 units of $2 water.
  4. Under the new rates, the rich pay 4*$2+2*$4 = $16 for water. The poor pay 3*$2 = $6 for water. Variable revenue is $22, but cost of delivery is 9*$4 = $36. Fixed charges need to rise by $14 to cover that gap. Since customers pay fixed charges per meter, each one pays $7 MORE. So now the rich pay $33 (less than before to consume the same amount of water) and the poor pay $17+$6 = $23 (greater than $18 from before). 
  5. Sure they are using more water, but the new tariff structure means that extra unit is costing them $5 instead of $2.
This argument makes one pause to consider the "fairness" of IBRs that not only increase consumption (!) but also shift more of the financing burden to the poorer (or lower water consumption) household.

I noted this problem in my book [3], but spent more time trying to design an IBR scheme (Some for Free, Pay for More) that would be more effective -- on the assumption that IBRs were here to stay.

Although I don't agree with everything Griffin and Mjelde say [4], this paper pushed me to reconsider IBRs; they seem to be more about engineering "proper" behavior than recovering costs or limiting demand [5].

At a minimum, I think we need to stop using IBRs. They are complicated to set up, may not be fair and may not even be efficient. It makes more sense to use URs (many "IBR" tariff structures are defacto URs structures) that are high enough to encourage conservation. In fact, I think it's worth looking into covering ALL costs with one tariff (no fixed charge per month), except that method would exacerbate the mismatch between the cost and revenue structure.

Note that the oil industry has VERY LARGE fixed costs but sells gasoline with URs. Water managers also prefer them. Half of the 308 cities in a recent global survey of tariffs (post coming soon!) used URs. Most of the others used IBRs (of which at least some were "de facto" URs).

I guess that my preferred alternative would have a fixed charge to cover fixed costs plus a UR to cover variable costs AND reflect water scarcity. Refunds of excess revenue (per meter, not per person) would transfer money from heavy to light water using households.

What about human rights, social equity and per person allocations? Those may be important in developing countries, but they are irrelevant in developed countries where $1 per 1,000 liters (250 gallons) of drinking water is still pretty damn cheap.

What do you think?

Bottom Line: IBRs do not deliver benefits in proportion to their complications. Better to charge a uniform price for water (as in a market) and promote conservation via a single, simple lever: higher prices.

Notes:
  1. The volumetric price per unit rises as you use more with IBRs; it does not with URs. RM sent this example of "innovative rate designs" as an example of the conventional wisdom.
  2. Since they face the same price "on the margin."
  3. See note on page 242 ("They prefer IBRs...") as well as Young and McColl [DOC]
  4. Their assumptions drive their results (it's possible to design "progressive" IBRs by increasing LATER blocks); their discussion of utility and social welfare does not match my experience in the water sector.
  5. See this post questioning their very existence.