By Eric de Place of the Sightline Institute

The cheaper the power, the more we use.

I'm going to geek out for a second. But first, check out this graph:

I suppose there are two lessons:

1. Price and consumption are not perfectly correlated. Clearly there are many non-price factors affecting electricity consumption. (These include, at least, the local climate, building size and type, and local energy efficiency policies.) But still...

2. Price definitely affects use, and the fit gets better as you move up the price axis. The more expensive electricity is, the less likely consumers are to be profligate.

In energy circles it's sometimes alleged that consumers are price insensitive or economically irrational about consumption. There's some truth to that, but it's only a partial truth.

These charts help demonstrate why carbon pricing can be effective. Putting a price on carbon -- or a price on energy -- acts to reduce consumption. Price is not the only factor and it may not even be the biggest factor, but it does appear to matter. And it appears to matter more above about 10 or 12 cents per kilowatt hour.

This hooks into a larger debate in the Western Climate Initiative.

Many utilities are arguing that they should get free carbon allowances on the grounds that they are rate-regulated, and so they will not (or cannot) take windfall profits. They even deploy arguments about consumer protection, pointing out that they can protect ratepayers by not passing on price increases. That sounds pretty good at first, but it doesn't work so well in practice. (For the definitive rebuttal to these arguments, see Clark's excellent post.)

The point is, for carbon policy to work fairly, the price of energy needs to go up to stimulate appropriate conservation. When the price of electricity is low, as it is in Seattle for example, even the most aggressive efficiency and conservation programs yield modest results. (And Seattle truly does have industry-leading programs.) Residents of California's air-conditioned cities consume less electricity, on average, than residents of either Seattle or Portland.

Luckily, under smart versions of cap and trade, higher energy prices can be good for consumers -- as counter-intuitive as that sounds. Here's how: WCI should auction carbon permits and allow utilities to pass on the prices to consumers. Then -- and this is critical -- WCI should refund the money to citizens, especially those below median income. These refunds (two flavors here and here) can easily make most energy consumers whole again. With cap and rebate, the "price signal" of carbon gets communicated through to consumers, but with no net pain for most.

(A footnote: there might still be inequities in a system like the one I've described. If you're a Seattle City Light customer, for example, your hydro-based utility needs to purchase very few carbon permits and so there would be only a very modest price increase at most. That would mean a continuation of the limited price incentive to conserve. But your neighbors in western Washington would see higher rate increases, and your friends in, say, southern California would see much higher increases. We might need to iron out these geographic inequities.)

Want more evidence of the relationship? Okay, then. Here's a chart for the 200 largest utilities in the United States. You see very much the same price-consumption trend as you see in the West. And the correlation appears to get stronger as prices get higher.

The data for these charts come from energy economics guru Jim Lazar. Jim also provided original versions of these charts and then I re-designed them to make them a little easier to read. The conclusions in this blog post are, of course, mine and not necessarily Jim's.

I certainly agree that the price of power has a significant effect on usage - and perhaps nowhere has this been better demonstrated than in successful shift in incentives for utilities in California from selling more power, to "selling" more energy efficiency (in CA we spend more per kWh than almost any other state, but our per capita energy bills are not bigger, largely because we use the same amount of electricity per capita now as we did in 1970... whereas the nation as a whole has increased per capita power consumption by more than 50% since then).

However, the connection between price and consumption is not shown convincingly by the plots you've posted. The R-squared statistics indicate what proportion of the variability is explained by the variable on the X-axis (in this case, price per kWh). In the first, only a quarter of the variability is explained (R-squared=0.26), and in the second, which you say offers "more evidence" of the connection, barely 17% of the variability is explained by price, which I would say qualifies as "less evidence". The correlations are positive in both cases, but hardly striking.

In order to see the price-consumption relationship clearly, you would need to normalize the datasets for things like the climate and economic health of the regions being examined. I don't doubt that you would see a strong correlation in such a normalized dataset - but you can't use the above plots to make a strong statement about the price-consumption relationship.

For more on R-squared, see: http://en.wikipedia.org/wiki/R-squared

Posted by: Zane Selvans on July 28, 2008 12:17 PM

The first graph that you have is a graph of kWh/yr v cost/kWh. At first blush it would seem logical that more expensive electricity would result in less consumption. But then I take a step back and think about why my consumption is low and it doesn’t have anything to do with price, leading me to ask if there is causality. I live in an area which has some of the highest costs for electricity in the nation (Manhattan – 22 cents per kWh) but my low consumption levels are defiantly the result of factors other than price. The first is the cost per square foot of real estate. With costs in the 2,000 – 5,000 per square foot range, it simply doesn’t make economic sense to have an apartment that is close to the same size as the average residence in Tacoma or Eugene. As such I don’t have space for many of the common loads, so I simply export those loads to local businesses (which I assume are not included in the Y axis). I don’t have space for a second fridge, so I store my beer, milk, soda at the fridge in the deli down the street effectively exporting that common load. I don’t have room for a washer/drier, so I export that load to the local dry cleaner. Kitchen is small and so I generally export the load associated with cooking and dish washing to one of the local restaurants. When the wife and I want to watch different shows I don’t have a second family room to go and watch my own shows, so I export that load to one of the local bars. If the electrical consumption associated with these loads was included in my annual usage (instead of the local businesses) I’m sure that my usage would start to deviate from the look of this graph. Living in a high rise building with one exposed wall (as most in Manhattan have), the R-value of my walls has to be significantly higher than the R-value of the average residence in Tacoma or Eugene. If our CDDs and HDDs are similar, my heating and cooling loads will be significantly less per residence (than the average residence in Tacoma or Eugene) but have zero correlation to price.

Given that I don’t think that the correlation in the price v usage graph is caused by price, I don’t think that the graph is an effective means to demonstrate why carbon pricing can be effective to reduce consumption. I do think that there should be carbon legislation. But I think it is a very far stretch to say that there is any causality demonstrated in the graph. It could be a very interesting world if you could assume causality based solely on a correlation. If so I’m going to start throwing a Superbowl champ parade at the beginning of each season for the Packers, cause there is a 1-1 relationship between Superbowl parades and Superbowl championships.

Posted by: snooj on July 29, 2008 12:46 PM

Please note that comments will remain open for only 14 days after the article is posted. While previous comments will remain visible, attempts to post new comments after this period will fail. This helps stop comment spam, so your forebearance is appreciated.

The Worldchanging comments are meant to be used for further exploration and evaluation of the ideas covered in our posts. Please note that, while constructive disagreement is fine, insults and abuse are not, and will result in the comment being deleted and a likely ban from commenting. We will also delete at will and without warning comments we believe are designed to disrupt a conversation rather than contribute to it. In short, we'll kill troll posts.

Finally, please note that comments which simply repost copyrighted works or commercial messages will be summarily deleted.