Utility and diminishing marginal utility

The traditional model assumes consumers are rational utility maximisers: they have consistent preferences, full information, and choose the bundle of goods that gives the greatest utility (satisfaction) from a limited income.

• Total utility (TU) — the total satisfaction from consuming a given quantity.
• Marginal utility (MU) — the extra utility from one more unit: MU = ΔTU ÷ ΔQ.

The first cold drink on a hot day is bliss; the fourth is barely worth drinking. TU is still rising, but by less each time — so MU is falling. If MU turns negative, TU actually falls.

Why this matters: diminishing marginal utility is the reason the demand curve slopes downwards. If each extra unit is worth less to you, you will only buy it at a lower price.

Your turn — marginal utility

The utility-maximising rule

A rational consumer spends a limited income so that the last pound spent on each good yields the same marginal utility. Otherwise they could reshuffle spending and gain.

Notice the assumptions: the consumer must know every MU, every price, and be able to do the arithmetic instantly. Stated baldly like that, the model is obviously false — which is exactly the behavioural economists' point.

Your turn — utility per pound

The utility-maximising rule

Imperfect and asymmetric information

Rational choice needs perfect information. In reality:

• Imperfect information — nobody knows every price, quality or long-run consequence. Buying a pension at 22 requires forecasting 45 years ahead.
• Asymmetric information — one side of a transaction knows more than the other. The used-car seller knows the car is a lemon; the buyer does not (Akerlof). The patient cannot judge whether the operation is needed; the surgeon can.

Consequences: adverse selection (only bad risks buy the insurance, so premiums rise, so good risks leave) and moral hazard (once insured, people take more risk). Both are forms of market failure — you meet them again in 4.1.8 and 4.2.4.

Policy response: compulsory information (energy labels, calorie counts, MOT certificates), regulation of professionals, and compulsory insurance to defeat adverse selection.

Bounded rationality and bounded self-control

Herbert Simon argued that people are boundedly rational: they want to choose well but face limits on information, time and cognitive capacity. So instead of maximising, they satisfice — they take the first option that is good enough.

Bounded self-control: even when people know the best action, they fail to carry it out. They over-value the present (present bias / hyperbolic discounting), so they under-save, over-eat and put off the gym. The rational model has no room for this — a rational agent who knows smoking kills simply does not smoke.

• Under-saving for retirement is bounded self-control, not ignorance.
• Choosing the same lunch every day is a rule of thumb (heuristic) — a cheap short cut that avoids the cost of full calculation.

Careful: boundedly rational is not the same as irrational. People are doing their best under real constraints — which is why predictable, systematic biases appear.

How firms use the same psychology

Governments nudge — but firms have been doing it commercially for far longer.

• Decoy pricing: a cinema offers small popcorn £3, large £7. Add a pointless medium at £6.50 and sales of the large jump — the decoy makes £7 look like a bargain by anchoring.
• Charm pricing: £9.99 is processed as nine-something, not ten. It exploits a rule of thumb (reading left to right).
• Free trials with automatic renewal: the default plus inertia does the selling.
• Scarcity cues: only 2 left at this price triggers loss aversion.

Evaluation: if firms can exploit biases, the market outcome is not welfare-maximising even with competition — a behavioural case for consumer protection regulation. Counter-argument: consumers learn, and heavy regulation of pricing raises costs and reduces choice.

Maximising or satisficing?

Biases and rules of thumb

• Anchoring — the first number you see drags your valuation towards it. A was £80, now £45 sticker makes £45 feel cheap, whatever the good is really worth.
• Availability bias — judging probability by how easily an example springs to mind. People overestimate the risk of plane crashes and shark attacks because they are vivid and reported.
• Loss aversion (Kahneman & Tversky) — a loss hurts roughly twice as much as an equivalent gain pleases. This produces the endowment effect: people demand more to give up a mug than they would pay to buy it.
• Social norms — behaviour copies the group. HMRC letters saying 9 out of 10 people in your area have already paid measurably raised on-time tax payment.
• Altruism and fairness — in the ultimatum game, people reject unfair offers even though rejecting leaves them with nothing. Pure self-interest cannot explain this.

The key claim: these errors are systematic and predictable, not random. Random errors would cancel out and the rational model would still work on average. Systematic ones do not — and that is what makes them exploitable, by firms and by governments.

Name that bias

Tap a behaviour, then tap the bias that best explains it.

Choice architecture and nudge

If presentation changes behaviour, then how choices are laid out is itself a policy lever. Thaler and Sunstein call this choice architecture, and a nudge is any change to it that alters behaviour predictably without banning options or significantly changing prices.

• Default choice — the most powerful nudge. UK pension auto-enrolment (opt-out, not opt-in) lifted workplace pension participation dramatically, because inertia now works for saving.
• Framing — 90% fat-free sells better than 10% fat, although they are the same thing.
• Restricted choice — offering fewer options avoids choice overload.
• Mandated choice — forcing an active yes/no (e.g. organ donation) so that inertia cannot decide for you.

Evaluate nudges: cheap and non-coercive, and they preserve freedom of choice (libertarian paternalism). But effects can be small and short-lived, they may not tackle the underlying cause (poverty, addiction), and there is an ethical objection — who decides which way to nudge? A tax or a ban is a shove, not a nudge.

Why is auto-enrolment a nudge?

Match the example to the concept

Tap an example on the left, then the concept it illustrates.

Your turn — total from marginal utility

What behavioural economics actually claims

Is the rational model dead?

Not quite. The rational-choice model survives because it is simple, general and often right.

• For: it predicts well in repeated, high-stakes, well-informed markets — commodity traders, experienced firms. Demand really does fall when price rises. And behavioural findings are often from small lab experiments that may not scale.
• Against: it predicts badly for one-off, complex, emotional or long-horizon decisions — pensions, mortgages, health, addiction. Exactly the decisions where the cost of a mistake is highest.

The mature position (and a strong AO4 line): behavioural economics does not replace the rational model — it tells you when to trust it, and it adds a cheap new policy tool alongside taxes, subsidies and regulation.

Exam tip: in a market-failure essay, always ask whether a nudge could achieve the same result as a tax at lower cost — and whether it is likely to be strong enough on its own.

The big ideas to know

You have covered the whole of AQA 4.1.2. Press Finish to see your score.