Lessons from Behavioral Economics for Rationalizing Energy Use*
This paper is based upon a presentation by its author at the 4th IAEE Asian Conference in Beijing.Roman J. ZytekSenior Economist, Middle East and Central Asia Department
International Monetary Fund(Washington, DC)email@example.com
Theoretically, households and firms choose the volume of utilities they use to maximize their welfare and profits, respectively. Government policies often try to influence household and firm choices to support specific social goals (Tanaka 2011). In the past, some policies encouraged more use. More recently, policies have started to encourage reduction in use, or more rational use. In most of these cases, policies aim to reduce the intensity of use, often at certain times of day. Policies to achieve these objectives involve changing prices as well as non-price measures.
Many energy-rich countries are considering reforms to reduce the rate of growth of their domestic energy and water use. The rapid growth in energy and water consumption has been, to a large extent, the result of low domestic prices for most utilities. So far, policymakers have focused on non-price measures to reduce demand growth. In the future, these measures will need to be supplemented by price reforms.
The increase in domestic energy and water prices will require deep reductions in energy intensity of the economy to preserve living standards and the economic competitiveness of domestic enterprises. The history of price reforms indicates that their success hinges less upon the policymakers’ ability to launch the reforms than on their ability to help consumers and firms change their habits, that is, learn to adjust their demand to changing market-based prices. Research shows that habits, such as utilities consumption, are entrenched; changing them takes time and costly effort. When planning price reforms in the utilities sector, it is therefore important to identify measures that may have been overlooked in the past to ensure that utility price liberalizations bring the desired change in habits among households and in the production sector as quickly as possible so as to help bring about the broader objectives of micro- and macro-economic adjustment.
This paper argues that price reforms, including liberalization, need to be reinforced with structural reforms that bring the quantity, price, and cost information on utilities’ use directly to the consumer’s attention. Price adjustments alone are not as effective as they could be because in the reality of today’s markets consumers have limited access to the relevant information needed to allow them to rationalize use, in particular in response to a sudden price change. The traditional explanations for low price elasticity of demand in the utilities sector mostly ignore the specific realities of how the information on unit prices, quantities, and total costs of utilities’ used is passed to users. As a result, only households and firms that actively seek information are likely to respond quickly to price changes to rationalize utilities’ use.
The Special Features of the Utilities Market
The term “utilities,” used to describe the services and their suppliers, has always been associated with necessities, such as water, electricity, natural gas, and telecommunications.1
Most utilities are not transacted in free, efficient markets. Unlike buyers in the farmers’ market or a grocery store, utility buyers have little idea of the volume and value of energy or water they consume, at least contemporaneously. In some markets, such as centralized home heating or water in multifamily buildings, and almost universally in television and telecommunications, consumers are not even allowed to truly select what they want. They can either buy an entire bundle of services (say, 50, or 200 TV channels; 300 or 500 telephone call minutes) or nothing. In some instances, the option of not buying does not even exist, because the services are offered in multi-utility packages, such as all-utilities-inclusive apartment rentals or bundled cable television, phone, and Internet access.
There is a strong historical justification for the current market structure in the utilities sector.Utilities have traditionally been delivered to households by government-owned or government-regulated monopolies. Most utilities were established as monopolies and most governments may have been keen to spread the benefits of electrification, municipal water and sewage, and telecommunications as broadly as possible. Increasing supply to increase the penetration of the services was the first priority. Limiting demand was not an issue. The delivery model favored lowest delivery cost and cross-subsidization of users. This model ensured the widest demand for the utility and helped reduce average unit costs by spreading large fixed costs common in the utilities sector among the largest possible number of users. Governments may have assumed that individual user demand for utilities was somewhat inelastic. The demand was restricted by home size, number of inhabitants, and the limited number of electricity- and gas-powered appliances. In many countries even today utilities are not sold to individual households; electricity, natural gas, water and sewage, home municipal heating, and trash removal services are still sold to neighborhood associations or entire apartment buildings.2
The utility meters were installed to measure total utility consumption, not to help consumers make informed purchase decisions. Most utility meters are installed either in basements, outside of the home, or in tightly locked boxes, and are not designed to provide much useful information to consumers. Most meters report quantity used, not the monetary cost charged to the user. In some cases, such as sewage or trash collection, quantities are estimated as proxies based on other indicators (such as water use or type of residence) or projected and averaged across a large number of buyers, not individually measured at the point of purchase. With the exception of retail gasoline, most households and firms use energy and water for weeks or months before seeing and paying any bill.3 As a result, households typically have little idea of how much electricity, natural gas, water and sewage, and telecommunication services they use. Moreover, they have a limited idea of how much these utilities will cost when they decide to use them.
Figure 1. Gas, electricity, and water meters in the author’s home.
The electricity meter is hidden under the stairs; the gas meter is placed on the side of the building; the water meter is placed in the basement. All three report quantities of the service used: kWh of electricity, cubic meters of gas, and liters of water. The information panels on all three meters are well covered, if not locked.
The historical organization of the utilities’ retail markets has lost most of its merit today, at least in the industrialized and middle-income countries. The bulk of today’s household demand for power, water, and other utilities does not come from necessities anymore. Basic hygiene needs, cooking and cleaning, localized heating and cooling account for a small share of utility demand. Instead, most of demand in advanced and, increasingly, middle-income countries comes from discretionary demand to operate increasingly powerful home appliances that are used to support leisure (ease daily household chores) and high-end entertainment. This trend of growing demand for utilities is likely to continue. The rate of growth may have slowed in the most advanced countries, but could re-accelerate as the world transitions to mass-market electricity-powered transport.4 Also, growth in electricity demand may accelerate if new time-saving technologies expand the range of available energy services.5 However, the outdated structure has become an obstacle to rationalizing demand for the old utilities and, increasingly, for the new ones, such as broadband, where policymakers try preserving incentives for providers to invest in new capacity while keeping consumer prices low in a market built around net neutrality and “all you can use” access at a fixed monthly fee.
Notwithstanding these developments, the organization of the utilities markets has remained essentially unchanged. So far, reforms in most countries have focused on creating some semblance of competition for some utilities, improving bill collections, and increasing communications between households and power companies to better match generation with demand through the use of smart meters. The more recent “smart metering” reforms are intended to integrate users into modern smart grids, with the main objective of improving matching electricity generation with demand. However, most of the efforts to improve collections or the introduction of smart metering are not designed to improve consumer decision making. Even today, most utility meters are installed far from customers’ views and do not supply consumers with useful information that could help them make better decisions about usage.
The following section first discusses how the existing utility metering arrangements limit users’ ability to rationalize consumption. It then proceeds to suggest specific measures to address the shortcomings of the existing arrangements.
Behavioral Economics Lessons for Reforming Utilities Markets
A. Lessons for Utilities from Behavioral Research
Behavioral research has identified major reasons for the slow response of utility users to price changes and suggested effective interventions to reinforce price signals. The main intervention is to bring the price and cost information to the user, and bring it at the right time, in the right place, and in the right form. Such intervention is not a substitute for genuine price reforms, but is a prerequisite, a necessary condition, to bring the benefits of price reforms to consumers, utilities, and the entire economy as rapidly as possible, and thus allow for effective adjustment to new prices.
First and foremost, utility users must be aware of prices and costs to respond to the price information (Hanna, Mullainathan, and Schwartzstein 2012). Real-life case studies and laboratory experiments find that people do not necessarily notice their energy use or costs. At the company level, the chief accountant may see a surge in energy expenses, but may have trouble convincing line workers to reduce energy use even if much of it is pure waste. Wasteful energy or water use may have been rationally ignored, or even condoned, when the utilities were provided at low cost. As a result, users are unlikely to have developed the needed behavior and supporting habits to search for price and use information to make rational choices when prices are raised. Jassoe and Rapson (2012), Faruqui and Sergici (2010), and Faruqui, Sergici, and Sharif (2009) show how bringing electricity use and costs to the consumer’s attention can make them rationalize their energy purchases and make significant reductions in use. These studies stress the importance of ensuring that households get rapid feedback on how much energy, and its cost, they use in affecting behavior; specifically, encouraging electricity conservation.
Second, the information must be current and relevant. Bushnell and Mansur (2005) find that people respond to their energy bills, not to current pricing, which they often do not know. Specifically, if their utility bills come with delay, their behavior adapts to past, not current or future prices. The information must be relevant. People must see the monetary cost of the energy used, not the amount of kilowatt hours (kWh) of electricity or British thermal units (Btu) of natural gas that flows through their home meters. Also, the data on kilowatt-hour use could be of little value when users pay progressive tariffs. Using a grocery store analogy, buying utilities in most countries is like going to a grocery store to pick up items without ever seeing prices and costs, and then being asked to pay the bill for a month or quarter’s worth of grocery shopping several months later.
Third, understanding how people respond to price information is critical to designing proper tariff structures.People seem to respond to average, not marginal prices (Ito 2012). As a result, electricity and water tariff reforms that implemented two-part tariffs to allocate fixed costs of building power generation to larger potential users (Fernàndez-Villadangos 2006), or progressive tariffs to discourage outright the use of electricity or water, have been quite inefficient (Borenstein 2010).6 Such reforms may create confusion among the public without bringing the desired behavioral change. Even if they are successful in reducing demand, they bring deadweight losses by discouraging socially and economically justified energy use. Also, they shift the demand for energy from efficient energy, such as electricity, to inefficient, but cheaper energy, such as from fuel oil, kerosene, or even burning household waste. They may have the unintended consequence of shifting the cost burden of electricity and water use from smaller, often more affluent families to larger, poorer families, an outcome that may become even more pronounced with the deployment of distributed renewable energies such as solar or wind. Finally, progressive tariffs may frontload demand early in the accounting period (calendar month) at the expense of possible significant underconsumption late in the accounting period, in particular when prepaid meters are used. As numerous studies of the paycheck cycle have demonstrated, cash- and credit-constrained consumers are often unable to spread their expenditure evenly over the pay period.
Fourth, the effort required to indentify and understand utility prices and the costs associated with using particular appliances must be low. Most people do not understand the different energy-related costs and potential savings from alternative items, such as home insulation, home appliances, or cars. Finding out and understanding the costs and potential benefits is too costly to be justified in the case of infrequent purchases of products in an environment of rapidly changing relative prices and technologies. This cost includes both the financial cost and time spent to find the information about a product’s energy intensity. Therefore, policymakers and manufactures need to develop simple and credible ways of informing consumers about real costs and benefits of their products. So far, the energy savings from many new products could be too small to give consumers an incentive to do intensive research before making purchases (Sallee 2013). Also, the endowment effect, and related loss aversion, and mental accounting may further explain why people do not make financially optimal choices when it comes to purchasing household appliances and cars.
B. Specific Measures to Help Utility Users Rationalize Consumption
Reforms of subsidies in the utilities sector, including energy subsidies, need to consider lessons from both economics and behavioral research. First and foremost, households and firms must see their utility costs, and see them contemporaneously with use, not months after the fact. Therefore, policies need to ensure that:
The recommended metering reforms could strengthen the effectiveness and cost efficiency of strategies aimed at developing good energy conservation habits and support the transition to dynamic, time variant retail tariffs. Strategies focused on developing good habits have been shown to be among the few strategies that are effective in having a lasting impact on human behavior, including energy use. For example, Hamamoto (2013) studies the cost of 14 relatively simple energy- and water-saving measures encouraged by the Energy Conservation Center, Japan (ECCJ) that could supplement the benefits afforded by using low-energy-intensive appliances. Though higher prices are needed to encourage saving behavior, Hamamoto and the studies summarized by Delmas, Fischlein, and Asensio (2013) stress that non-price measures are needed to change habits. Some of the measures to improve habits include periodic and recurring energy audits, energy saving contests, and electronic reminders. Electricity, gas, and water meters providing the right information, in the right place, and at the right time could go a long way in reducing the costs of the various non-price interventions aimed at energy and water conservation. Improving habits takes time and effort, a lot of both, and can be very labor intensive, especially in the absence of supporting technology that provides the needed information.
So far, the utilities companies have mostly ignored lessons from psychology and behavioral economics in their efforts to convince the public to accept an efficiency enhancing innovation. The attempts to implement dynamic pricing meters and contracts have faced stiff opposition from consumers (Albright 2010). The utilities companies’ support for dynamic pricing is viewed by consumers with suspicion. The suspicion is warranted. Without the implementation of proper metering, as defined above, dynamic pricing may just increase prices, in particular through price discrimination, and transfer ever larger consumer surpluses to the utilities sector, without giving households and businesses a chance to rationalize behavior.
One question that needs to be addressed is how to pay for the reform. Ensuring that meters in new establishments and replacements of old meters deliver the right information in the right place and time will be important. Also, installation of new meters could be economically justified even before the old meters are fully depreciated whenever the savings in utility costs from the new meters are larger than the cost of the new meters and their installation. In some situations the metering reforms could be close to self-financing. Specifically, if new meters are installed in individual housing or business units and replace community meters, the reform may generate rapid reductions in utility use and reduce the risk of future volatility in use. These two benefits could instantly increase the market value of the housing or business units implementing the reforms.
Conclusions and a Word of Caution
The role of the utilities sector has dramatically changed over the past sixty or so years. The sector has moved from the state of infancy to full maturity. It started as a tightly regulated supplier of small doses of highly beneficial necessities. Today, it is a full-fledged supplier of wholesale quantities of services that satisfy increasingly higher-end discretionary needs. Yet many segments of the utilities sector continue to operate, price their services, bill and collect payments as if they were still old-style providers of indispensable necessities. As a result, utilities continue to be sold in inefficient markets that limit competition and consumer choice and implicitly encourage overuse and overspending. Reforms to increase the sector’s efficiency by bringing in the private sector have not much changed the structure of the markets. Unfortunately, given the rigidities and entrenched position of the incumbent companies, regulators may need to step in to tilt the balance of power in the relationship between sellers and buyers towards the middle.
Policymakers can strive to rationalize utility consumption by reinforcing the sound price mechanisms with non-price strategies.7 Policymakers should complement price reforms with measures that bring the price and cost information right to the buyer-consumer in an accessible form, content, and place, and at a time when such information can influence consumer buying decisions. The paper offers intuitively simple measures that accomplish this task. They extend to the utilities sector practices that are already accepted in markets for most goods and services. They require minimal regulatory intervention: only the development of effective metering and disclosure standards. Similar interventions have long been recognized as effective and desirable in other markets.
Concluding, a word of caution is needed. Behavioral research has documented that people who lead efforts to change behavior suffer, like everyone else, from optimism bias. They have a strong and very natural tendency to systematically over-project the benefits of the projects and regulations they support and legislate (Tal and Cohen-Blankshtain 2011). Moreover, results from laboratory and limited field studies have shown to be difficult to generalize (Al-Ubaydli and List 2013). Therefore, any complex policy efforts should be carefully assessed and balanced against their direct, indirect, and opportunity costs, all measured over short-, medium-, and long-term horizons.
* The views expressed in this paper are those of the author(s) and do not necessarily represent the views of the IMF, its Executive Board, or IMF management.
 In Arlington, Virginia, condominium buildings completed as recently as the mid-1990s were still equipped with water and sewage metering for the entire building only.
 Most post-paid utilities’ bills are sent monthly or quarterly.
 See Bashmakov (2007) for a comprehensive discussion of “energy transition.”
 See Keay (2007) for a discussion of the historical drivers of energy demand and possible nature and sources of demand for energy services in the future.
 Borenstein (2010) discuses why progressive tariffs are socially suboptimal. They result in economic deadweight loss. Replacing flat and progressive tariffs with dynamic, time-varying tariffs, intended to balance demand with supply at a moment’s notice and reduce demand for generating capacity, are encouraged instead (Allcott 2012; Borenstein 2005, 2012; Faruqui and Hledik 2009; and Faruqui, Hledik, and Tsoukalis 2009).
 See Madrian (2014) for the most up-to-date discussion of how policymakers around the world are increasingly turning to behavioral economics for ideas when designing policies; and Allcott (2014) for the latest review and discussion of public policies aimed at increasing energy efficiency that draw on lessons from behavioral economics.
Al-Ubaydli, Omar and John A. List (2013). “On the Generalizability of Experimental Results in Economics: With A Response to Camerer.” NBER Working Paper No. 19666.
Albright, Julie M. (2010). “Customer Behavior Project (USC) -DOE/ LADWP Smart Grid Demonstration.” USC Viterbi, School of Engineering, Information Sciences institute, available at: http://csis.org/files/attachments/110331_Energyja.pdf.
Allcott, Hunt (2012). “The Smart Grid, Entry, and Imperfect Competition in Electricity Markets.” NBER Working Paper No. 18071.
Allcott, Hunt (2014). “Paternalism and Energy Efficiency: An Overview.” NBER Working Paper No. 20363.
Bashmakov, Igor (2007). “Three Laws of Energy Transitions.” Energy Policy 35 (July): 3583–3594.
Borenstein, Severin (2005). “Wealth Transfers from Implementing Real-Time Retail Electricity Pricing.” NBER Working Paper No. 11594.
Borenstein, Severin (2010). “The Redistributional Impact of Non-linear Electricity Pricing.” NBER Working Paper No. 15822.
Borenstein, Severin, 2012, “Effective and Equitable Adoption of Opt-In Residential Dynamic Electricity Pricing.” NBER Working Paper No. 18037.
Bushnell, James and Erin T. Mansur (2005). “Consumption under Noisy Price Signals: A Study of Electricity Retail Rate Deregulation in San Diego.” Journal of Industrial Economics 53(4): 493-513.
Delmas, Magali A., Miriam Fischlein, and Omar I. Asensio (2013). “Information Strategies and Energy Conservation Behavior: A Meta-Analysis Of Experimental Studies From 1975 to 2012.” Energy Policy Vol. 61 (October): 729-739.
Faruqui, Ahmad and Ryan M. Hledik (2009). “Transitioning to Dynamic Pricing.” The Brattle Group. Available at SSRN: http://ssrn.com/abstract=1336726.
Faruqui, Ahmad, Ryan M. Hledik, and John Tsoukalis (2009). “The Power of Dynamic Pricing.” The Brattle Group. Available at SSRN: http://ssrn.com/abstract=1340594.
Faruqui, Ahmad and Sanem Sergici (2010). “Household Response to Dynamic Pricing of Electricity - A Survey of the Empirical Evidence.” available at SSRN: http://ssrn.com/abstract=1134132.
Faruqui, Ahmad, Sanem Sergici, and Ahmed Sharif (2009). “The Impact of Informational Feedback on Energy Consumption -- A Survey of the Experimental Evidence.” available at SSRN: http://ssrn.com/abstract=1407701.
Fernàndez-Villadangos, Laura (2006). “Are Two-Part Tariffs Efficient When Consumers Plan Ahead?: An Empirical Study.” CREAP2006-04. Centre De Referència en Economia Aplicada. Electronic copy available at: http://ssrn.com/abstract=1852258.
Hamamoto, Mitsutsugu (2013). “Energy-saving Behavior and Marginal Abatement Cost for Household CO2 Emissions.” Energy Policy 63 (December): 809-813.
Hanna, Rema, Sendhil Mullainathan, and Joshua Schwartzstein (2012). “Learning Through Noticing: Theory and Experimental Evidence in Farming.” NBER Working Paper No. 18401.
Ito, Koichiro (2012). “Do Consumers Respond to Marginal or Average Price? Evidence from Nonlinear Electricity Pricing.” NBER Working Paper No. 18533.
Jassoe, Katrina and David Rapson (2012). “Knowledge is (Less) Power: Experimental Evidence from Residential Energy Use.” NBER Working Paper No. 18344.
Keay, Malcolm (2007). “Energy: The Long View.” Oxford Institute for Energy Studies, SP20.
Madrian, Brigitte C. (2014). “Applying Insights from Behavioral Economics to Policy Design.” NBER Working Paper No. 20318.
Sallee, James M. (2013). “Rational Inattention and Energy Efficiency.” NBER Working Paper No. 19545.
Tal, Gil, Galit Cohen-Blankshtain (2011). “Understanding the Role of the Forecast-maker in 0verestimation Forecasts of Policy Impacts: The Case of Travel Demand Management Policies.” Transportation Research Part A: Policy and Practice, Volume 45(5): 389-400.
Tanaka, Kanako (2011). “Review of Policies and Measures for Energy Efficiency in Industry Sector.” Energy Policy 39 (October): 6523-6550.
Volume 23, Number 1 Announcements