3.1.2.4 - Behavioural economics and policy

Choice architecture and framing

The way alternatives are arranged and described can influence decisions even when their underlying facts are unchanged.

Choice architecture is the framework in which people encounter a choice: the ordering of options, their visibility, the steps required to select them and the information displayed. A cafeteria can place fruit at eye level and make it easy to pick up, while still offering the same desserts at the same prices. Limited attention and habitual routines make the location matter.

Framing is the way a decision or information is presented. An illustrative product has a 95% chance of working for one year and a 5% chance of failing during that year. “95 in 100 work for a year” and “5 in 100 fail within a year” express the same probabilities. The first emphasises a favourable outcome; the second makes failure prominent. Different choices in response to the two frames can show that presentation, not only objective facts, affects the decision.

Framing is not permission to distort facts. Showing only a monthly charge while omitting a compulsory fee creates an information problem. A fair comparison should preserve relevant conditions and denominators. “Save £10” says little without the original total cost or the resources needed to obtain the saving.

Governments and other agencies can use architecture to make helpful choices easier, but businesses can also use it to encourage profitable purchases. The designer's objective may differ from the consumer's. Transparent information, easy comparison and meaningful access to alternatives therefore matter when judging the design, not just whether it changes behaviour.

Nudges and behavioural mechanisms

A nudge encourages a choice through its presentation or ease while preserving the person's ability to choose an alternative.

A nudge works through a behavioural mechanism, usually without a substantial change in financial incentives. Moving water to the most convenient position in a canteen can reduce the attention or effort needed to choose it. A timely reminder to renew a discount travel card can address inattention. An accurate message about neighbours' energy conservation can use a social norm. The explanation should identify the obstacle and how the intervention changes it.

A tax on sugary drinks changes relative financial costs. A ban removes an option. Both can affect behaviour, but they operate differently from the choice-preserving presentation changes used in these nudge examples. A policy may combine a nudge with such measures; calling the entire package a nudge obscures which mechanism produced the outcome.

Suppose a library sends an easy-to-read reminder shortly before a book is due. Forgetful borrowers may return books sooner, increasing availability for other readers. The reminder does not make a late return impossible or impose a new fine. It may have little effect on someone who deliberately keeps the book because it is still needed, so the relevant obstacle matters.

Nudges are not automatically cheap, effective or beneficial. Designing messages, reaching users and maintaining systems costs resources. A reminder can arrive at an inconvenient time; a norm message can legitimise undesirable behaviour if it advertises how common it is. The policy should be judged against a clear objective and evidence of the response, while retaining genuine ease of choosing otherwise.

Default choices

A default determines what happens when a person takes no active step to choose another option.

A default choice is a preselected option that the individual can change. In an illustrative workplace savings scheme, new employees are enrolled unless they opt out. In an opt-in design, no saving occurs unless employees enrol. The contribution and all available alternatives can be identical; what changes is the consequence of inaction.

Defaults can matter through inertia, the effort of changing a setting and a belief that the preselected option is recommended. They may help someone who intends to save but keeps postponing enrolment. They can also enrol someone whose circumstances make the default a poor choice, so clear information and an easy opt-out are important.

An invented randomised trial compares two designs with identical contributions and information:

The difference between the rates is $70-40=30$ percentage points, not 30%. As a percentage of the initial 40% rate, the increase is $\frac{70-40}{40}\times100=75\%$. The 120 additional savers are evidence that the default changes participation in this trial. Random allocation and unchanged contributions make a default effect plausible; the table does not establish that every extra saver is better off or will continue saving for years.

A government or employer should consider contribution affordability and the option to exit as well as enrolment. A profitable subscription renewal default may exploit inattention and lead to unwanted purchases. The behavioural mechanism can be similar even though the welfare consequences differ.

Restricted and mandated choice

Limiting available options and requiring an active decision change the choice process in different ways.

Restricted choice limits the number of alternatives. A savings provider could offer a short list of clearly differentiated funds rather than hundreds. This may reduce choice overload and make comparison manageable. However, the restricted menu can exclude a fund better suited to a particular person's needs. A restriction can also remove a harmful option; its mechanism includes changing what is available, not merely making one option more noticeable.

Mandated choice requires the individual to make a decision, rather than allowing inaction to select a default. For example, an illustrative legal rule could require a person to select “share my data” or “do not share my data” before completing an application. It compels an active decision, while allowing either answer. This can reveal preferences more clearly and reduce reliance on a designer's default. It also requires attention and may create frustration or rushed choices when decisions are complicated.

AQA's terminology describes a required decision, usually required by law. It should not be confused with a rule that requires one particular outcome, such as compulsory purchase of a specified product. The latter removes the freedom to choose a different outcome. AQA subject-specific vocabulary, 2024, p. 4.

Consider three settings with the same two data-sharing options. A preselected “do not share” setting is a default; showing only “do not share” removes an alternative and restricts choice; requiring an explicit selection between the two is mandated active choice. Which design is preferable depends on the risk of harm, how well people can understand the decision and the importance attached to individual autonomy.

Evidence and policy judgement

A useful policy judgement links the behavioural mechanism to measured outcomes, costs, distribution and a clear account of whose welfare matters.

An illustrative council wants households to waste less water. It considers a truthful norm message saying that comparable neighbours conserve water, alongside a simple reminder near the tap. These address social influence and inattention. They may work less well when the household already wants to conserve but has a leaking pipe: a message cannot repair that constraint. Matching the intervention to the cause of behaviour is the first step.

To test the policy, randomly assign comparable households to receive the message or no message and measure water use over the same period. Compare actual usage, rather than relying only on stated intentions. A common weather change affects both groups; a comparison helps avoid crediting the message for that change. The trial still needs enough observations and a check for long-term persistence, and a local result may not transfer to households with different circumstances.

A fall in water use is only one outcome. It may save resources, but the council should weigh the cost of the programme and whether essential use falls for vulnerable households. Ask whether people understand the message and retain easy alternatives. “The government should encourage conservation” includes a value judgement about priorities; the measured response helps inform that judgement but cannot determine the moral weight of freedom, fairness and environmental protection by itself.

A defensible conditional conclusion is that a transparent, inexpensive nudge is promising when inattention or norms are important and a credible trial shows a lasting beneficial response. It is less suitable as the sole policy where physical constraints or serious harms need another response. Behavioural insights extend the policy toolkit; they do not remove the need to compare costs, evidence and preferences.

Quick recap

Architecture and framing shape how choices are encountered. Nudges preserve alternatives; defaults choose the outcome of inaction. Restricted choice reduces alternatives, while mandated choice requires an active decision. Policy effectiveness and welfare must be judged from the mechanism, credible evidence, costs and distribution, not participation alone.