09 November 2011

Another comment on TEoA

Jeffrey Ripp writes:
I finished reading your book (and your short piece on utility insurance) ... last week. I appreciate the effort and thought you put forward in these pieces; I have spent many hours thinking about the themes you write about in the book, and I wanted to share my thoughts.

I am intrigued by the utility insurance concept, but I am not exactly sure what the companies would be insuring or how this would improve customer service. Would it be a catastrophic failure such as a main break or a boil-water order due to some contaminant in the system? I think that customers do in fact understand the value they get from a private water system – they expect water that meets SDWA requirements (safe to drink) 24/7. But I agree that they don’t understand the work or effort to get it to their homes – some of the outputs are not obvious. Also, customer bills don’t necessarily reflect the increased costs right away – the cost of main repairs are spread over a number of years under the test year/regulatory model. This isn’t to say that insurance wouldn’t work, but I am not exactly sure how this would help to attract capital to the utilities or prevent over-investment. I agree with the importance of benchmarking performance – we are in fact trying to do this in Wisconsin. We need to improve our measures to move away from outputs to outcomes, but we have started this effort at least. You can get a sample on our website.
My reply:
Insurance -- like car insurance -- would help clarify the difference between skill and hard work (via lower premia) and bad luck. Breaks, accidents will still happen but the rates would reflect whether they are accidents or not. Compare this to the current norm of accident = "rates go up" even if we don't know WHY the accident happened.

Second, insurance payouts would keep rates from rising after ACCIDENTS. They would rise after INCOMPETENCE.

Insurance would help financiers understand the quality of the management team, and thus the risk their money faces.

Overinvestment would lower premia, but they would go TOO low (no risk of accident even), thereby attracting attention.

Benchmarks -- as data -- are good. People will interpret the data in different ways but they NEED data to interpret. That's why I am starting the water data hub in 2012.
Jeffrey replied with: One more point to clarify – under the insurance model. If understand this correctly, if a utility has made proper investment and demonstrates performance at or above benchmark levels (top 25%), but there is a catastrophic main break, the insurance company would pay the costs to repair and replace this main? What about the situation in Milwaukee a few decades ago, where there was a crypto outbreak that lead to widespread sickness and death? The end result was a need for additional treatment. Assume that there had been no reason to assume that the water treatment that had been in place for 100 years was insufficient (which was the case). Would the insurance company pay for the necessary plant investment to install ozone systems for disinfection? I am just trying to get my head around exactly which risks the insurance would be covering and how this could work.

One thing to note is that main breaks happen all the time and don’t necessarily result in higher rates. Even utilities with adequate main replacement schedules can run into distribution system problems that are addressed as part of the O&M expenses, but do not necessarily go up dramatically after a single event.

DZ: The insurance would cover legitimate costs, but not costs due to negligence. (Another way of looking at it is as a warranty). The crypto outbreak would be insured in the sense that it was unexpected but NOT if Mil water had delayed installing an ozone system.... Insurance risks and coverage would have to be spelled out in detail, of course.

Sure, they do that via "self insurance" but we want to make sure that they are not OVER-insuring, i.e., charging customers for incompetence (as LADWP did with breaks related to its "pulse"watering schedule)

Additional comments:

pp 38-39 – IBRs. The points you raise are valid; however it is possible to design IBRs to target the additional expenses caused by peak demand using a cost of service allocation – this is how we attempt to do it in Wisconsin. The results are not perfect, but certainly, some customers get a better price signal. I agree that such rates must be carefully designed or they will fail to accomplish the policy goals.

DZ: Agreed. I have taken a stronger stance on IBRs recently, mainly to point out how URs may be easier to design and implement than IBRs. Higher prices are good for cutting demand in either case :)

pp.42-44 – Some for Free. I like this approach in concept, but there are some drawbacks. Since water utilities are so capital intensive, the fixed cost portion of the bill would be very large compared to the variable charges, diluting the price signal the customer sees (anti-conservation). This sounds a lot like the “decoupling” advocated by the water industry. Typically, we try to set the fixed portion of the residential bill between 25-50 percent of the total for an average customer, but if we were to allow the fixed charges to match the fixed charges as allocated through a cost of service study, they could be in excess of 75 percent of the total bill. While our approach does result in some revenue instability, it provides a better conservation price signal to the customer. As an aside, I don’t like the water budget model approach – it is overly burdensome, for one, and customers don’t usually like government or businesses collecting what they feel is “private” information. An analogy is that we don’t have to input the number of people in the car when we fill up at the pump, e.g., driving a van with 8 people vs. an SUV with 1 person. To me, the question is not how people use the water but WHEN and HOW MUCH? If the customer causes peak demand, these variable costs should be assigned that cost-causer. I am not sure that budget rate does that.

DZ: We agree on water budgets. Look at that chapter again. I am basically saying charge 75% Fixed to cover fixed costs and ANOTHER 75% (revenue = 150% of costs) on variable costs. Such overcharging sends a conservation signal without destabilizing revenues. I am NOT in favor of decoupling (blog post to come, also to appear in 2nd edition of TEoA) b/c it says "do what you want, we take your money," which makes little sense to customers. Also note that decoupling -- like IBRs -- does not happen in free markets. Bad sign.

Jeffrey replied with: One of the challenges (not insurmountable) that we have in a regulated utility environment is that we are constrained by the model used for rate base/rate of return ratemaking. We cannot exceed the cost of service in designing rates. So it is difficult to price externalities through our rate designs - 100% of costs is the limit. Unless we come up with a consistent and equitable way to do so, I am not sure what kind of a “premium” could be placed on volumetric rates to exceed the variable portion of costs determined in a rate case. Maybe it will come through carbon taxes/carbon pricing tied to electricity used for pumping, maybe there will be some intrinsic public value assigned to the water resource, but this will certainly vary by watershed.

DZ: Oh, sorry. The excess 50% charge is rebated to customers, to keep revenue equal to costs. Another way of seeing it is as a charge for "reliability" which is a justifiable cost.

pp. 47-48 – bill frequency and format ARE overlooked. These are critical to sending the price signal you want. Plus, municipal bills are usually complicated by non-utility charges (wastewater, stormwater, solid waste, fire protection, etc…) The water bill may only account for 25 percent of the total.

DZ: Yes.

pp. 93-94 – Economic Develop Rates are cropping up in Wisconsin – I am not sure that our Commission understands the policy implications here, but you point them out well. I am intrigued by the IBR for business approach – I have been advocating for a demand-based cost of service approach for several years, rather than our traditional customer class definitions of industrial, residential, commercial. We already do this for energy pricing, and this idea makes sense to me for water. I would like to pilot this idea for a Wisconsin utility.

DZ: Glad you like the idea. Let's talk more nuts and bolts as you try to develop it.

p. 168 – Hookup Fees/Revenue – I agree that the housing market trends masked the underlying revenue trends, and now that the market is weak, we are seeing the effects of demand reduction. I have been saying this for years, but nobody believes me.

DZ: Maybe they will believe US! :)

pp. 152-153. Benchmarks. I have been wondering if there is way to incentive performance using benchmarks. Maybe we could provide a higher rate of return for top performers? The insurance concept seems to assume that O&M costs drive rates, but this isn’t always true. We must factor in capital costs (depreciation, return on investment, and property taxes or equivalent).

DZ: Agreed. Insurance needs to be "calibrated" to these. Rate of return for superior performance will come via lower insurance premia, which can be passed directly into profits (for IOUs) or management bonuses for muni managers/staff.

p. 173. Global water crisis. I agree absolutely – there is none. Unlike energy, water is local. There may be common themes, but the solutions are all local.

DZ: Yep. Need a bumpersticker for this.

p. 200 Water Markets. Some interesting concepts here, but I think the ability to implement is low, due to institutional barriers and data limitations. For example, how do you deal with existing groundwater withdrawals, which are not easily measured? What about new entries into the market as people are born – how do you allocate new water rights to them? What about people that move from Wisconsin to California? Do they forgo their rights in WI and gain rights in California? How do you overcome the transaction costs for trades? What about natural variability – who loses rights when the water is not available? I am not discounting this altogether, but I see this as a difficult thing to implement even within a single watershed.

DZ: Groundwater needs to be monitored and perhaps regulated, everywhere. I handled newborns by allocating tradable water according to the population (that changes every year). A local rights market would change as people move in/out of the state, but it's better to base it on a watershed (or nationally). Transaction costs can be VERY high, but "decision rules" and automation can reduce them, e.g., to one or two sessions per year for someone like you or me who can specify goals. Water rights fluctuate with annual flows.

All of these questions are useful, but my big idea here is to escape conventional ways of thinking about water rights -- and ESPECIALLY to find a way to get income from "national wealth" to poorer people in LDCs. Doesn't mean I don't want to see this happen on the Colorado!

2 comments:

  1. Enjoyed the back and forth and now that my head has stopped spinning i should probably read the book:)

    But I'm mostly intrigued by this:
    "That's why I am starting the water data hub in 2012"

    What's the plan?

    ReplyDelete
  2. @Cary -- http://waterdatahub.org/

    ReplyDelete

Spammers, don't bother. I delete spam.