Monday, July 23, 2012

Elasticity isn't rigid... (with update)

It's been a while since I've posted (again) and I promise to get better about regular activity.  However, some things have gotten my attention lately.

At a recent meeting, I heard someone in a position of responsibility over electric rates repeat the old saw that “electric demand does not vary with price.”  To use the economic-egghead terminology “electricity is price-inelastic.”  I was proud of myself for only convulsing mildly. 

I am here to tell you that this is flat-out wrong. 

“Electricity is not price elastic” is an oft-repeated old saw, not because it is true, but because it is an oft-repeated old saw.

Let me explain:  When those responsible for determining electric rates and tariffs believe that electricity is price inelastic, they price it as though it is price inelastic. 

As a result, the customer never sees a price that reflects the existence of elasticity, so the customer never acts in a way that reflects price elasticity.  So the observer says, “See? No elasticity!”

It is a tautology.  It isn't seen as price elastic because retail rates are designed with the assumption that prices are inelastic, so customers do not see or, respond to, price changes.

Price elasticity is not an either-or.  Commodities are more or less price elastic in relation to other commodities.  Let me say that again.  Price elasticity is measured in relation to other commodities

To say that one commodity "has no price elasticity” because it is less price elastic than other commodities is to apply an absolute reference to a relative measure.

Put another way, to say that electricity is “not price elastic” because it's less price elastic than say, footballs, or books, or luxury goods (which can be inversely price elastic, where demand rises with price and perceived exclusivity) is to treat a continuum as a binary. 

Yes, electricity is less price elastic than some other commodities (and perhaps more elastic than others) but that does not mean that electricity is insufficiently price elastic to provide demand response.

Let's take a look at whether real-world prices might change enough to create some elasticity.  A quick scan of PJM's LMP prices (available at shows some useful numbers.  Pretty much at random, I pulled the “Historical OVEC Pricing Point LMPs” for November 1, 2009.  These are hourly Locational Marginal Prices, expressed as both a real time price and as a Day Ahead hourly bid/offer.

Minimum RTLMP (Real Time Locational Marginal Price)-24.92
Maximum RTLMP (Real Time Locational Marginal Price)292.54
Average RTLMP (Real Time Locational Marginal Price)35.37

Now, that negative number means that they'll pay you to draw power during that hour, which seems counter-intuitive, but it really happens.  Overall, the price swing (low to high) is 9 times the average price, and the maximum price is 8 times the average.

Okay, that's just too much price swing to comprehend, and a negative price is too much of a mind-bender for you.  Let's look at the next column, the day-ahead LMP, which is more stable:

Minimum DALMP (Day Ahead Locational Marginal Price)4.58
Maximum DALMP (Day Ahead Locational Marginal Price)107.96
Average DALMP (Day Ahead Locational Marginal Price)35.21

The maximum price is 23 times the minimum, and about 4 times the average.  Anything is demand elastic at that kind of price swing.  If the price of your next new car varied that much based on how you bought, you can bet that people would change how they buy cars.

“Customers don't want to pay that much for electricity!” I hear.  To that I respond “But they already do.”  Unless the PUC is setting confiscatory rates (that don't cover the cost of purchased power) then the rates that are paid for every kW/h reflect some kind of composite price that includes, but to some extent hides, those price swings. 

“People will refuse to pay that much for power!”  Yup, they will refuse.  They will find ways to not take power at those prices, and take power when it's cheaper.

That's called elasticity of demand.

I'm not advocating an overnight change here.  We are looking at a transition in how people use and think about energy in order to take advantage of that real-world price elasticity.  Of course, I've said that before.

But the longer we hold on to tautological thinking about the relation of price and demand, the longer that transition to a more realistic and efficient electricity market will take.

UPDATE 7/25:

I've heard some feedback on this blog post that indicates that people believe that I'm anti-subsidy. I'm not. Subsidies allow systems to serve important public goals. What I'm against is implicit subsidies, like those buried in the traditional way we price electricity at retail.

Implicit subsidies are sloppy. They aren't trackable, and they often get left subsidizing things that don't need to be subsidized (or subsidize things that are contrary to the public interest, requiring an even larger subsidy in the opposite direction). Read one of my earlier posts for an example.

Remove implicit subsidies and, to the extent possible, price energy at the real cost of providing it at the time it is provided. The closer you get to that, the better.

Then explicitly subsidize where there is a need.
  • You want to subsidize air conditioning for the elderly and infirm so they don't keel over in a heat wave? Great! I'm going to be elderly and infirm someday, if I live long enough (and I fully intend to, Lord willing). So create an explicit trackable subsidy to do that.
  • Do you want to subsidize the poor who don't come out better off under time of use rates (according to a Brattle Group study, all 20% of them)? Well, okay, but do it explicitly so that if (or when) it creates unexpected consequences, you can adjust it explicitly without breaking everything else.

Is it complicated? Yes. Is it more complicated than what exists now? No. All that changes is that the complexity becomes transparent, rather than hidden. Transparent systems are no more complex than implicit systems, and may be less so, because you can change one thing without affecting others as much.

It's a little like one of those fine swiss watches with the open framework. You look at it and the innards seem hopelessly complex. What you don't see is that a cheap watch has all the same components as a fine swiss watch, but you don't see them. The only difference is transparency.

No comments:

Post a Comment