Imperfect Information, Biased Reasoning, and the Energy Efficiency Gap

In a new working paper, David Austin argues that for a variety of reasons, private agents do not optimize with respect to energy efficiency measures. From Addressing Market Barriers to Energy Efficiency in Buildings:

… I argue that imperfections—relating to misperceived prices, imperfect information, and biased reasoning—in markets for energy-using products interfere with people’s ability to make privately optimal decisions. The result is an energy-efficiency gap, or an “energy paradox”: Energy-efficient technologies with lower lifetime costs diffuse more slowly through the economy than would be expected given their cost advantages.

Because of energy’s social costs—not only regional pollutants (primarily particulates and oxides of sulfur and nitrogen) and global greenhouse gases (primarily carbon dioxide) from energy production and consumption, but also local pollution, traffic, and noise from resource extraction and transport—there are social benefits from policies that narrow the energy-efficiency gap. Growing concerns about global climate change have made it important to identify ways of reducing greenhouse-gas emissions at relatively low cost. Policies that address imperfections in markets for energy efficiency can reduce polluting emissions at a cost that is relatively low compared with the benefits. With buildings responsible for a substantial share of U.S. energy consumption, such policies make buildings an important potential source of lower-cost emissions reductions.

Austin cites the three sources of these distortions:

  • Energy prices are misperceived and may differ from the incremental cost of service;
  • Consumers’ responses to price signals are hampered by imperfect information; and
  • Consumers’ assessments of potential energy savings tend to be too low because of biased reasoning.

The second reason was more familiar to me; underprovision of information is commonplace when the benefits cannot be appropriated. The first and third are (to me) more novel. He cites as an example the tenant/landlord issue, or “split incentive” problem. Builders/landlords are unable to recoup the higher costs associated with greater investment in energy efficiency since evidence of that investment for tenants is difficult to discern. The third reason is not one that is typically assumed in neoclassical economic analysis, but seems to be important in practice.

As a consequence, putting a price on carbon emissions reflecting current estimated damages would not eliminate the energy-efficiency gap (the difference between existing and economically-rational levels of energy efficiency), or the rationale for other energy-efficiency policies. But higher future carbon prices, reflecting rising present-value damages, would further narrow the gap. This point is illustrated, in a heuristic fashion, in Figure 1:


Figure 1: Energy-efficiency gap.

The stakes are large. Policies to reduce energy consumption in buildings by 5%, narrowing the energy-efficiency gap by half, would result in reduced environmental damages by between $600 million to $4.6 billion per year (in 2012 constant dollars).

The working paper is here.

13 thoughts on “Imperfect Information, Biased Reasoning, and the Energy Efficiency Gap

  1. tj

    Does this assume policy makers have perfect information? Seems like a stretch. One could argue that each 1% that consumers contribute to the ‘energy gap’ due to imperfect information is offset by policy makers’ overestimation of policy benefits due to imperfect information.
    For example, let’s assume that a landlord could paint his roof white to cool the atmosphere or leave it black so that global warming continues unabated. Policy makers then mandate that all rooftops should be painted white, because their current state of the art climate/economic models show white rooftops save energy, reduce carbon emissions and slow man-made global warming.
    However, 5 years after implementing the mandate, they notice huge costs to society from less rainfall, because their models did not inlcude the cost of unintended consequences. In this simple example, white rooftops reduce rainfall.
    I realize that my story is not perfect, but it demonstrates that policy makers make mistakes due to imperfect information as well, yet we trust them as experts.

  2. dilbert dogbert

    An engineer friend who worked for Chevron told me during the first energy crisis no efficiency improving project would be considered for equipment unless the stack temperature was near 900 degrees and the payback time was less than 3 years.

  3. jonathan

    I skimmed the paper and thought the author should spend some time with a developer or three. And some people in the trades.
    The real point of leverage is the financing. If lenders drive building toward energy efficiency, it will happen. If not, then it won’t because lending drives the spreadsheet. Building practices changed overnight when lenders put in covenants about asbestos. How to get energy efficiency into lending? That’s the question. It could be done through some sort of portfolio measure, meaning the bank would need to show some sort of “social good” to get some benefit – or face some cost. The justification would be the relationship of the bank to the government. The blunt reality is that if banks say, “You’ll get the money if …” then the “if” tends to happen.

  4. Lord

    Information costs can’t be underestimated, but apply equally to the situation to which these improvements have to be applied. Insulation may be great until you realize there is no attic and would only be possible through reroofing. Solar may sound nice until you find you use only a tenth of energy its economics is based on.
    Another complication is technological improvements may result in lower future costs by delaying long lived ones while technological unimprovements may lead to higher future energy costs, so there is no visible future, only distorted reflections of the past.

  5. Preston

    I’ve worked in the energy efficiency world, public and private, for nearly 20 years. This paper doesn’t really plow any new ground. It’s correct, but most just reiterates that same things that have been done for quite some time. Wisconsin has a very good energy efficiency program in Focus on Energy, which provides unbiased information and offers incentives for purchasing high efficiency equipment. The program works mainly through market channels, i.e. trade allies, big box stores, equipment vendors to reduce the chicken-egg issue of energy efficiency. (Q.Why didn’t you buy high efficiency? A. The store/vendor didn’t carry it. Q.Store owner/vendor, why don’t you carry efficient items? A. People don’t buy them, too expensive.) Energy efficiency program can help reduce the number of chicken-egg situations as well as build awareness of and sustain demand for high efficiency items. But the most cost effective way to move efficiency forward is via constantly updating standards and codes. Once the program has built acceptance and marketshare of a technology, society needs to make standard or code changes that lock in that level of efficiency. Then the program stops providing incentives for the new standard level technology and starts offering incentives for a new, more efficient technology if one that works can be developed. That in a nutshell is how programs have to work. So without a strong effort to move standards and codes ever upward (yes regulations), most people will continue to buy the lowest initial cost item time after time for a variety of reasons…which the paper doesn’t really get into. Information is the least effective and most unreliable means of saving energy.

  6. person1597

    Access, Benefit, Continuity — low information energy consumers take this for granted… until the lights go out.
    Consumer habits anchor the expectation these ABC’s are unassailable. Thanks to saint Nick [Tesla], data-flow lifestyles are a birthright! One Nation Under Grid!
    Our memories and experiences will guide our decisions,[Lord, above] fixing goals on fantasy solutions drawn from model based scenarios targeting politically impossible issue rectification timelines.
    How will we know if the solutions are working? What will the best asymmetrical information be at the time of measurement? Will we ever have perfect information, or just more controls over dissemination?
    Ultimately, how can our energy wastefulness be curbed?
    Isn’t it ironic that deflation is likely the most potent driver of energy conservation in the here and now…(?!)
    During deflationary episodes, overcapacity is dismantled, aggregate demand weakens and lower efficiency units are taken offline. Downsizing has a silver lining — less inefficient operation.
    Do we want to promote energy-efficiency through such market based adaptations? Sure, if the organic marginal demand can be balanced by upgrading efficiency as well as promoting new research and development.
    Rather than pinning insufficient capacity as our biggest social ill (thanks, Dick) we should first remove the inefficient capacity and then replace it with something better that addresses the long term issues like biotoxicity.
    Of course it is ridiculous to use economic controls to target consumption trends… (or is it?)… but what other tools are there besides coerced social reprogramming or macroeconomic chaos that otherwise changes habits?
    Government dictate?! Oh, gosh, more asymmetry.
    Or, just maybe going outside to feel that trillion-trillion watt bulb shining over our heads. Stop overeating! Exercise! Do nothing!
    I’m glad this blog paying attention to our evolutionarily novel, ubiquitous, unlimited energy-on-demand economy. We are totally dependent on this fragile status-quo for human existence.
    If losing ten pounds would help, I’m in for 30.
    Wow, like, the sun is shining in the PNW! See ya!

  7. Bruce Hall

    Energy efficiency comes in many shades and the decision information at one point may not be relevant … or true … during subsequent points.
    For example, in 2007 when natural gas was about $12-13 per kcm, what would the investment decision have been for upgrading a heating system? With that same commodity priced between $2-3 in 2012… 5 years later… what would the decision be?
    We live in a world of energy volatility. It seems the marketplace responds to out of balance energy equations with corrections. Hopefully, the millions of energy decisions that are made each year result in an overall rational process… even if individual decisions are made imperfectly.

  8. ezra abrams

    like medical care, a building is something you live with for a long time;
    New stuff often doesn’t work, or is over sold, or is not understood
    therefore, a prudent rule of thumb for life is never (unless it is an emergency) take a drug untill it has been on the market for a few years; don’t build with new technology.
    a few years ago, NJ was in a townhouse boom: suburban attached houses, basically a long long rectangle of a building.
    and builders put in a new type of roofing element.
    turns out, the new roofing wasn’t quite fire resistant, so whena fire started in one unit, it spread quickly to others…
    and older guys will remember the aluminum wire fiasco: when aluminum wire was first being use for ROMEX, to carry electricity around in your house, people didn’t know hat if you connect an aluminum wire to a switch with copper contacts, you have a problem…..

  9. valuethinker

    David Laibson at Harvard has done a lot of work on ‘hyperbolic discounting’. See also Richard Thaler at Chicago.
    Humans, irrationally, apply much higher discount rates to short term decisions than long term ones.
    Thus precommitment works in pension schemes. Give people the chance to ‘opt out’ not to ‘opt in’ and you get 65% joining the pension plan, not 35%.
    Households and companies typically demand 30-40% internal rates of return on energy efficiency investments (ie payback in 2-3 years).
    That’s irrational for what is essentially a risk free investment in an environment where US TIPS pay negative real returns, and nominal interest rates on mortgages are around 5%.
    It’s a private discount rate far in excess of any plausible social discount rate. Most of the energy efficiency literature assumes a discount rate of c. 10% in calculating NPV. To get a nuclear power plant to be economic you need to assume an interest rate of around 6-7%.
    Efficiency standards get around this problem, just as compulsorary pension contributions do the savings problem.
    There is a distributional effect at low income levels in that houses and appliances and cars cost more (but those costs of meeting the standards tend to fall over time: refridgerators are not only 70% more efficient than 30 years ago, they are also cheaper).
    The problem of hyperbolic discounting, and of information problems and ‘split incentives’ (particularly in commercial rental buildings, and private rentals) is so large that the best way to overcome these things is generally progressively improving standards, plus financial assistance for low income households to achieve these goals.
    I know these problems. Displays as we enter the building tell us kwhr pa (about 1.75m kwhr so far this year– enough electricity for 500 average homes). Yet all the halogen spotlights in the foyers bathrooms and elevators are still 50 watt halogens rather than 6 watt LEDs. Given many of these lights are on 24/7, payback on the LEDs which cost 5 times as much would be less than 9 months. However the landlord has no incentive (tenants pay the electricity bills) and no one tenant has responsibility for lighting in common areas.
    The UK is the poster child though on improving housing standards. UK housing is amongst the most poorly insulated in Europe. Programmes administered by the utility companies to improve energy efficiency amongst poor customers and over 65s have helped to reduce energy consumption per square foot by c. 30% since 1970.
    In addition some of the efficiency gains were eaten up in a rise in mean home temperature. From 13 degrees C (about 58) in January in 1970 to about 19.5 degrees C (about 70) now.
    That’s got significant health benefits, as studies show cold is one of the major reasons for morbidity and death amongst 65+ generation.

  10. valuethinker

    Just to pull out that last point… once you realize energy efficiency standards are a significant public health issue (houses that are cheaper to heat or cool are more likely to stay above the minimum temperatures for health, and below the maximum ones) it adds a whole new dimension to the situation.

  11. valuethinker

    dilbert dogbert
    c. 30% pa Internal Rates of Return (the discount rate that sets Net Present Value = 0 ie costs=benefits) are typical of energy savings targets in industry. ie paybacks of 2-4 years.
    Consumers will not generally invest in energy savings with lower than 20-30% IRRs.
    However preferences are not consistent. New windows have amongst the longest payback periods(20 years+ single pane to double pane, in an English climate). Because they have aesthetic issues, and because they are heavily *sold* by double glazing companies, they are among the most common energy efficiency improvements installed.
    A PVC U double glazed window of unexceptional energy performance is unlikely, in the UK, to pay back its cost before it wears out– they last about 25 years before brittleness and discoloration finishes them off (and leakage is a constant issue often forcing replacement long before that). PVC U is an environmental nightmare for safe disposal, due to heavy metal content– a lot probably just wind up in landfill.
    The reasons for their popularity as a home improvement are 1). aesthetic 2). lack of information (perhaps) and 3). the room does feel warmer (cold air convection currents created by warm air hitting the window and going down are reduced).
    The same logic (3) applies to going from double to triple glazing. The energy savings are not worth the cost in a UK climate (say average January temperature around 4 degrees C/ 40 degrees F) but the room does feel warmer.
    A related problem is the lack of post installation performance review. The standard now in the loft is to go from 100mm insulation between the joists to 270mm. Payback on that is usually less than 3 years *but* there is loss of storage space as the insulation now sits above the rafters.
    In addition, if the loft insulation is not well installed then air movements will negate much of the improvement due to cold bridging to the outside walls.
    I doubt most people notice the £100 pa drop in their heating bills from increased loft insulation– gets lost in the noise.
    Just like changing the lightbulbs for more energy efficiency versions, saving perhaps £5 or £10 pcm, gets missed.

Comments are closed.