Mind the Energy Efficiency Gap

The energy efficiency gap describes the difference between the socially optimal adoption of energy-efficient technologies and the adoption actually observed. This apparently simple definition conceals an important measurement problem: the “optimal” level is a counterfactual that depends on assumptions about costs, energy prices, and consumer heterogeneity. An efficient appliance may be worthwhile for a high-use household but not for one expecting to move soon. Accordingly, non-adoption should not automatically be treated as irrational. Nevertheless, even after accounting for information problems, credit constraints and transaction costs, adoption may remain below conventional predictions. Behavioural economics offers additional explanations that households may have limited capacity to process lifecycle costs and place disproportionate weight on immediate expenditure. They also tend to avoid perceived losses, neglect low-salience energy costs and learn from peers.

Behavioural frictions can create genuine private inefficiencies in particular settings, but policymakers ought to distinguish these from rational responses to uncertainty, heterogeneous preferences and unmeasured costs. Policies should then be matched to the diagnosed mechanism instead of treating all under-adoption as the same problem. Allcott and Greenstone similarly conclude that evidence for a pervasive gap is mixed and that welfare gains are likely to be greatest when interventions are carefully targeted.1

The Hidden Complexity of Simple Choices

The first friction is bounded rationality. Choosing between appliances requires a household to forecast energy usage, compare expected lifetime costs, and discount uncertain future savings. This is a demanding optimisation problem, especially because such purchases are infrequent and consumers have little opportunity to learn through repetition. The bounded rationality model suggests that people therefore use simplified decision rules rather than solving the complete optimisation problem.2 They may choose a familiar brand, accept a retailer’s recommendation or focus primarily on the purchase price. These shortcuts emerge because careful calculation requires time and effort. However, when high cognitive costs systematically direct buyers towards inefficient products, private welfare can fall.

The corresponding policy should simplify the decision rather than merely provide more information. Standardised labels can translate technical efficiency ratings into expected annual or lifetime monetary costs. Online calculators, trusted advisers and trained sales staff can further reduce the effort required to compare alternatives. Yet disclosure is not automatically effective. In a Norwegian retail experiment, Kallbekken et al. (2013) found no significant effect from lifetime-energy-cost information for fridge-freezers. For tumble dryers, combining the information with trained sales staff reduced the average energy consumption of products sold by approximately 4.9 per cent, although the effect weakened over time.3 Information must therefore be understandable, timely and actively communicated at the point of choice rather than placed passively on a label.

The Cost You Never See

A closely related friction is limited attention and salience. Energy use is largely invisible when it occurs and appears only later on a periodic bill. By contrast, an appliance’s sticker price is immediate, visible and easily compared. Consumers may consequently overweight quality/price ratios and underweight operating costs, even where both are financially important.4 Salience theory formalises this tendency by proposing that attributes which stand out receive disproportionate decision weight.

Policies can respond by displaying estimated running costs beside the purchase price, providing frequent consumption feedback and sending reminders when households are most able to act, for example shortly before a boiler replacement. Smart-meter displays and redesigned bills can make consumption and expected savings more noticeable. However, adding more numbers may itself increase cognitive burden. Effective disclosure should therefore be simple, personalised and restricted to the few quantities most relevant to the decision. Information may also need to be paired with a clear recommended action, such as an efficient replacement option or a pre-arranged energy assessment.

Today Always Looks Expensive

Present bias provides a distinct explanation for delayed investment. Under standard exponential discounting, a person applies a consistent discount rate to all future outcomes. A present-biased consumer instead gives extra weight to costs and benefits occurring immediately. Energy-efficient technologies have precisely the troublesome structure: a large and salient upfront payment produces a stream of smaller, uncertain savings in the future. A household may therefore intend to install insulation or replace an inefficient appliance “later”, but repeatedly postpone the action when payment becomes immediate. Quasi-hyperbolic discounting captures this behaviour by applying an additional discount factor to future outcomes, producing time-inconsistent choices.5

Policies targeted at present bias should reduce or reorganise the immediate cost. Upfront rebates are likely to attract more attention than equally valuable future tax credits, while low-friction financing can align repayments with the flow of energy savings. Automatic enrolment into retrofit schemes with an opt-out and commitment mechanisms may help convert intentions into action. Nevertheless, policymakers need to distinguish present bias from genuine liquidity constraints. A household may reject a positive-return investment because it cannot borrow, not because it undervalues the future. In practice, the two mechanisms can interact, supporting the combination of simplified financing with behavioural design rather than relying solely on either subsidies or information.

Am I Losing It?

Loss aversion and status quo bias create further inertia. Prospect theory proposes that people evaluate outcomes relative to a reference point and experience losses more strongly than equivalent gains. For energy efficiency, the payment and disruption associated with a new technology constitute certain losses relative to the current situation, whereas future bill savings are uncertain gains. Replacing a functioning boiler may therefore feel wasteful even when doing so has a positive expected return. Status-quo bias reinforces this tendency by making the existing technology the psychologically preferred option.6

Policy can alter the reference point by describing inefficient consumption as an ongoing loss. For example, it should clarify that a household is currently losing several hundred pounds each year, rather than presenting efficiency only as a possible future gain. Performance warranties, trusted certification and insurance against installation failure can lessen perceived downside risk. Besides, scrappage payments or early-retirement incentives may also overcome replacement-cycle inertia by compensating households. However, such framing must remain transparent, since exaggerated claims could manipulate consumers and undermine public trust. Moreover, encouraging premature replacement is not always socially beneficial once manufacturing emissions and disposal costs are considered.

Am I the Only One?

Social norms and peer effects demonstrate that decisions are not made in isolation. When performance is uncertain, observing neighbours’ solar panels, heat pumps or electric vehicles provides information regarding reliability and signals what is locally normal. Bollinger and Gillingham (2012) found that an additional installation increased the probability of subsequent adoption within the same area.7 This points to a social multiplier: demonstration projects, neighbourhood ambassadors, referral schemes and visible certification can spread practical knowledge as well as pro-efficiency norms.

Comparative home-energy reports similarly tell households how their consumption compares with that of similar neighbours. Allcott and Rogers (2014) suggest that the effects of repeatedly delivered social-comparison reports remained relatively persistent after the programme stopped, although they subsequently decayed by approximately 10 to 20 per cent per year.8 Norm-based policies nevertheless have a double edge. Informing an already efficient household that it consumes less than average can encourage consumption to rise towards the norm, a “boomerang” effect. Adding an approval signal for efficient households can help preserve their behaviour.9

Matching the Policy to the Problem

These frictions rarely operate independently, which explains why a single label or subsidy may have little effect. Evidence from the US Weatherization Assistance Program is instructive. Fowlie et al. (2018) found that actual savings were substantially below engineering projections: projected savings exceeded realised savings by more than three times, while investment costs were approximately twice the value of actual energy savings.10 This warns against assuming that every technically efficient measure must also produce positive private or social returns. A coherent policy package might instead combine a simple eligibility message, upfront finance and practical help with disruption. Removing hassle costs may be as important as increasing the nominal subsidy. This system-based approach embodies the interaction of information, behavioural biases, financial and institutional conditions.

Footnotes

  1. Journal of Economic Perspectives, Is there an energy efficiency gap? (opens in a new tab), 2012.

  2. The Quarterly Journal of Economics, A behavioral model of rational choice (opens in a new tab), 1st February 1955.

  3. Journal of Consumer Policy, Bridging the energy efficiency gap: A field experiment on lifetime energy costs and household appliances (opens in a new tab), 2nd October 2012.

  4. Journal of Political Economy, Salience and consumer choice (opens in a new tab), 1st October 2013.

  5. The Quarterly Journal of Economics, Golden eggs and hyperbolic discounting (opens in a new tab), 1st May 1997.

  6. Econometrica, Prospect theory: An analysis of decision under risk (opens in a new tab), March 1979Journal of Risk and Uncertainty, Status quo bias in decision making (opens in a new tab), March 1988.

  7. Marketing Science, Peer effects in the diffusion of solar photovoltaic panels (opens in a new tab), 20th September 2012.

  8. American Economic Review, The short-run and long-run effects of behavioral interventions: Experimental evidence from energy conservation (opens in a new tab), October 2014.

  9. Psychological Science, The constructive, destructive, and reconstructive power of social norms (opens in a new tab), May 2007.

  10. The Quarterly Journal of Economics, Do energy efficiency investments deliver? Evidence from the Weatherization Assistance Program (opens in a new tab), 29th January 2018.