Utility Pricing: Rate Re-Structuring and Price Elasticities

As a part of my Master’s Capstone project in Urban Planning, I worked for the California Governor’s Office of Planning and Research to provide recommendations on utility rate restructuring using Los Angeles Department of Water and Power customer electricity use data. To view the entire report please click here.

Executive Summary

Setting the right price for electricity is both a political and economic issue as it has widespread implications for both consumers and the electricity market (Reiss & White, 2005). One of California’s priorities when designing electricity rate payment structures (RPS) is to conserve energy while ensuring that the tariff is not regressive (Reiss & White, 2005). There have been many pricing strategies in California that attempt to encourage conservation while remaining equitable. Since 1980, utilities in California have used an Increasing-Block Pricing (IBP) or tiered rate structure for electricity (Picker, 2012). An IBP rate structure has different “tiers” of electricity consumption where each tier is charged more per Kilowatt hour (kWh) than the previous tier. Customers who use less electricity are charged less per kWh; customers who use more electricity fall within the next tier and are charged more per kWh.  The Los Angeles Department of Water and Power (LADWP) use a three-tiered rate structure in the high season months, June through September.

In this paper, I use customer-level data from LADWP to answer two questions:

  1. The first question to be answered is how pricing differences, represented by the zone boundary, effect electricity consumption.
  2. The next question is if customers change their electricity use based on the average price of electricity or the marginal (tiered) cost of electricity, and what are the resulting price elasticities.

Furthermore, this paper will address how LADWP’s current rate structure affects disadvantaged populations.[1]

LADWP created two “Zones” of pricing based on temperature, where Zone 2 is a warmer zone and Zone 1 is the cooler zone. In efforts to equitably price all of their customers, LADWP gave a larger allocation of kWhs in each tier to Zone 2 customers to compensate for the warmer weather in the high season. Zone 1 has increments of 350 kWh per tier and Zone 2 has increments of 500 kWh per tier. Thus, consumption should be similar on average, across zones. This price difference only occurs during the high season; in all other months pricing is flat.

This study uses the households within a one-mile buffer of the two LADWP Zones of pricing. Selecting customers on the boundary of Zone 1 and Zone 2 controls for changes in weather, building characteristics, and demographics. Therefore, the affect price has on electricity consumption can be isolated.  A random sample of customers within the Zone boundary was taken from December 2011 to June 2012.

To answer how the different price schedules affect consumption, I ran five regressions, each model getting progressively more stringent with each iteration. The model estimates electricity use in the high and low season for customers in each Zone and for non-lifeline and lifeline customers within each Zone, controlling for household demographics, weather, and time variation.

Two key findings result: there are two unintended consequences of the pricing strategy. The lower average price per kWh in Zone 2 during the high season results in Zone 2 customers consuming significantly more electricity per customer, about 15% more than Zone 1 customers. The pricing strategy in Zone 1 curtails consumption for both non-lifeline and lifeline customers in the high season. In Zone 1, non-lifeline customers increase their consumption only 0.01% from the low season to the high season. On the other hand, Zone 2 non-lifeline customers increase consumption 15%. Thus, the pricing strategy is not improving the welfare of Zone 1 customers and not achieving its goal of reducing consumption in the high season.

Secondly, the pricing strategy has an unintended consequence within a Zone. Lifeline customers, who are by definition the lowest users of electricity, disproportionately reflect observed curtailment in high season consumption. The models estimate that lifeline customers in Zone 1 are actually decreasing consumption from the low to high season, whereas in Zone 2 it remains the same. The pricing instrument is therefore having an unintended effect in Zone 1, in that those customers who consume the least—lifeline customers—are absorbing a disproportionate amount of the curtailment of electricity use.

Seven important findings from this work are listed below:


Central Finding

There are heterogeneous effects by zone and lifeline status.

Finding 1

Zone 2 customers consume 6.5% (64 kWhs) more electricity in the high season compared to Zone 1 customers.

Finding 2

Zone 2 customers increased consumption 15% in the high season.

Finding 3

Zone 1 customers decreased consumption by 0.01% in the high season.

Finding 4

Across both seasons and Zones, lifeline customers consume 77 kWhs less than non-lifeline customers.

Finding 5

Zone 1 lifeline customers decrease consumption in the high season.

Finding 6

The majority of curtailment from the pricing mechanism is deriving from lifeline customers.


The next question I endeavored to answer is if customers change their electricity use based on the average price of electricity or the marginal (tiered) cost of electricity. The average cost of electricity meaning the average cost per kWh each customer pays and the marginal cost meaning the cost per kWh in each of the three tiers. Three regression models were run estimating the price elasticities of electricity demand based on the average cost and marginal cost, and also providing elasticities by lifeline status. The price elasticity using average cost is estimated at -0.519. The estimated price elasticity using marginal cost is more inelastic at  -0.242. This indicates that customers change their consumption behavior based more on the average price per kWh than the marginal cost per kWh, though both average and marginal cost cause significant consumption changes.

However, we see that lifeline customers are more inelastic. The change in the electricity use is less than for non-lifeline customers, as their price elasticity is -0.137.  Lifeline populations also do not change their electricity use based on the average cost of electricity, but the marginal cost. The interaction between lifeline customers and marginal cost is significant with a price elasticity of  -0.122. This indicates that lifeline customers display different behavior than non-lifeline customers, they are more inelastic to price changes and they also react more to marginal rather than average cost.

The elasticities estimated by average and marginal cost fit well within the normal range of elasticities found in California. The major findings are listed below:

Summary of Findings

Price Elasticities

Finding 7

The price elasticity using average cost is -0.560.

Finding 8

The price elasticity using marginal cost is -0.229.

Finding 9

Thus, customers react more to the average cost of electricity than the marginal cost, though both significantly change consumption.

Finding 10

Lifeline customers react to the marginal cost of electricity, not to the average cost.

Finding 11

Lifeline customers’ price elasticity of demand is -0.122. This indicates that lifeline customers display different behavior than non-lifeline customers, they are more inelastic to price changes and they also react more to marginal rather than average cost.

Finding 12

The elasticities estimated by average and marginal cost fit well within the normal range of elasticities found in California.


There are many policy implications for energy planning, utility companies, electricity pricing, and equity as a result of the findings highlighted in this work.

Recommendation 1: There are barriers to understanding the marginal price schedule which prevents customers from behaving rationally to the increased charge in every tier. Understanding the tiered schedule would require customers to invest time into researching the complex structure of utility pricing schedules. Utilities could work to educate their customer base about pricing schedules, and also make efforts to make it easy and simple to understand. Additionally, users could be provided with real-time information on how much they are consuming every day in order to make rational choices about when to conserve.

Recommendation 2: The main finding of this study is that there are unintended consequences within Zones and unintended consequences between Zones of the pricing strategy. As we saw, the price of electricity has a significant effect on consumption, even when controlling for variation in weather patterns across the city. The policy implications that arise out of this analysis depend on the goals of electricity providers and energy planners in the state.  On the one hand, the goal of having tiered pricing during the high season is to manage consumption so that the electricity grid is not overly strained. On the other hand, it is expected that electricity consumption will increase in the high season since temperatures are higher and more energy is required to keep buildings cool. The two different pricing schemes in the two Zones achieve one or the other of these goals. If the goal is to keep electricity use consistent in both low and high seasons, then the price is set right in Zone 1. However, if the goal is to allow for a reasonable increase in consumption during the high season, then the allocation of kWhs in Zone 2 is correct.

Recommendation 3: The pricing scheme’s objective is to induce conservation in the high season. The unintended consequence of this is that the majority of this conservation is from lifeline customers, who are the lowest users of electricity. Greater conservation could be achieved if this curtailment was focused on those customers consuming the most electricity, rather than the least. The objective is to get a larger fraction of non-lifeline customers to conserve during the high season.

There are different methods to achieve this. The most straightforward method is to increase the price. As prices increase, higher-end users will become more elastic to price. This too has unintended consequences as it creates greater disparity between lifeline and non-lifeline customers. Secondly, there could a uniform increase for all tiers.  The caveat being that the very lowest users does not have to participate in the pricing structure. In theory, the most vulnerable populations at the lower-end of consumption will not have to curtail consumption as much, while at the same time the increase in price will get more curtailment from the high-end users. However, this could put a squeeze on the middle-income class.

A limitation of the study is that the elasticities and differences in consumption applies to customers within the one mile boundary into each Zone; in as far as the overall sample is representative of the overall customer class, these results apply.

[1] Disadvantaged populations are here defined as any population in which LADWP has established an electricity “lifeline” program. These populations include customers that are low-income, elderly, disabled, on life-support or have a life-threatening disease

Below is a visual representation I created of the results found in this study:

Client Project UCLA


One comment

  1. Pingback: ABOUT ME | Jacki Murdock

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s