Why standard theory falls short

The standard microeconomic model assumes rational, time-consistent, narrowly self-interested, fully informed agents. These assumptions are useful — they produce clean comparative-statics predictions and a coherent welfare analysis. They are also wrong in patterned ways. Behavioural economics is the systematic attempt to repair the model with empirically-validated psychology while preserving as much analytical machinery as possible.

The four standard-model assumptions and how they fail

1. Rationality (preference axioms)

Standard model assumes completeness, transitivity, and choice-consistency. Empirical reality:

• Intransitive choices — people prefer A to B, B to C, but C to A in repeated experiments
• Choice-reversal — preferences depend on how alternatives are framed or grouped
• Menu-dependence — adding an irrelevant option changes the choice between original options (the Decoy Effect)
• Default-sensitive choices — the default option is selected far more often than non-default alternatives, even when defaults are clearly arbitrary

2. Time consistency

Standard model assumes exponential discounting. Empirically, people use quasi-hyperbolic discounting (covered in Micro module 8) — sharper discounting for the near future than the far future. This produces time-inconsistent choices: today's plan for tomorrow gets overridden tomorrow.

3. Self-interest

Standard model assumes narrow self-interest. Empirically:

• Fairness concerns — people pay to punish unfair behaviour (Ultimatum Game), tip waiters they'll never see again, donate to charity
• Reciprocity — kind acts elicit kind responses, hostile acts hostile responses, even when not in the agent's narrow interest
• Social comparison — utility depends on relative as well as absolute outcomes
• Identity — choices express identity (occupation, consumption, brand) in ways that aren't just consumption utility

4. Full information

Standard model assumes agents have all relevant information and process it optimally. Empirical reality: agents are subject to cognitive biases (availability, representativeness, anchoring) and have limited information-processing capacity. Even with full information available, attention and salience matter.

System 1 and System 2

Daniel Kahneman (2011, Thinking, Fast and Slow) popularised the dual-process model:

• System 1 — automatic, fast, intuitive, effortless. Operates by pattern-matching and heuristics. Always running
• System 2 — deliberate, slow, calculating, effortful. Engaged only when System 1 fails or the decision is judged important enough

Most economic decisions in daily life are made by System 1. The standard economic model implicitly assumes System 2 throughout — which is correct only for important decisions where the agent has incentive and time to think carefully. Most consumption choices, default selections, snap judgments are System 1 — vulnerable to all the biases and heuristics System 1 employs.

Prospect theory

Kahneman and Tversky's (1979) alternative to expected-utility theory. Three key innovations:

• Reference-dependent — value is measured relative to a reference point (typically status quo), not in absolute wealth levels. Gains and losses are felt relative to the reference
• Loss aversion — the disvalue of a loss is greater than the value of an equivalent gain. People feel a $100 loss roughly twice as keenly as they feel a $100 gain. The kink in the value function at the reference point
• Probability weighting — small probabilities are over-weighted; moderate to large probabilities are under-weighted. People over-pay for lottery tickets (small probability of large gain) and under-insure against small-probability large losses

Status quo bias

A direct consequence of loss aversion + reference-dependence. Once an option is the default or status quo, switching to an alternative involves treating the alternative as a 'gain' (relative to status quo) AND treating the status quo as a 'loss' (relative to the alternative). The asymmetric weighting means even slightly-better alternatives don't get chosen.

Examples in African contexts:

• Pension contributions — the default contribution rate captures most participants. Raising the default raises contributions; few people adjust the default downward
• Bank-account defaults — once a customer opens an account at a bank, switching to another bank is rare even when the new bank offers better rates
• Mobile-money provider choice — Safaricom's M-Pesa user base is sticky in ways that the rational-choice model can't fully explain
• Crop choice — farmers who plant maize one year mostly plant maize the next, even when alternative crops have higher expected returns

When standard theory still works

Behavioural economics doesn't mean abandoning standard theory. It adds patches to a useful framework. Standard theory works well when:

• Stakes are high — major life decisions (home purchase, retirement planning, choice of career) often engage System 2 deliberation
• Frequency is high — agents who make a similar decision repeatedly learn the optimal pattern (financial traders, professional poker players, expert chess players)
• Information is good — when agents have access to relevant information and time to process it
• Competition selects for it — in competitive markets, firms that don't optimise are competed away. Most product-market behaviour by firms is closer to rational than household consumption behaviour

Standard theory works poorly when:

• Stakes are low — micro-choices (which product to grab from a shelf, which default to select) are quick and behavioural
• Frequency is low — rare decisions don't generate learning
• Information is poor or biased — agents can't fully evaluate options
• Self-control is required — when the choice involves trading off short-term pain for long-term gain (saving, exercise, learning)

The behavioural approach to policy

Two contrasting policy stances:

• Paternalism — agents make mistakes; government should override their choices. Mandatory pension contributions, prohibition of certain products, tax cigarettes
• Libertarian paternalism (Thaler-Sunstein) — design choice architectures that nudge agents toward better choices without restricting their freedom to choose otherwise. Default options, framing, disclosure

Most modern behavioural-policy applications are libertarian-paternalist. Examples: pension auto-enrolment (UK 2012); organ-donor default opt-out (most European countries); calorie labelling at restaurants; sugar-tax structured as a price signal rather than a prohibition.