Behavioral Economics
The Death of Homo Economicus

1 · Introduction

For two centuries, economic theory rested on a useful fiction: that agents possess stable preferences, complete information, unbounded computation, and unwavering self-control. They maximize expected utility. They update beliefs by Bayes' rule. They are, in the language of Richard Thaler, Econs.[1]

The trouble is that Econs do not exist. Real humans — Humans, in Thaler's bifurcation — anchor on irrelevant numbers, fear losses twice as much as they enjoy equivalent gains, succumb to defaults, and cannot remember to save without a payroll deduction.

Behavioral economics is the project of replacing the simplifying assumptions with empirically grounded ones. Its method is the laboratory experiment, supplemented by field trials, neuroimaging, and increasingly, large-scale A/B tests in industry.

2 · System 1 and System 2

Daniel Kahneman's Thinking, Fast and Slow (2011) crystallized a framing developed across cognitive psychology. The mind operates two modes:

System 1 is fast, automatic, intuitive, emotional. It identifies a friend's face, completes "bread and ___," reacts to a thrown ball.

System 2 is slow, deliberate, effortful, logical. It computes 17 × 24, parks a car in a tight spot, follows a complex argument.

System 1 is always running. System 2 is lazy and expensive. When System 1 produces a confident answer, System 2 typically endorses it. Most behavioral failures are cases where System 1's quick heuristic is wrong and System 2 fails to intervene.

3 · Prospect Theory

Kahneman & Tversky (1979) replaced expected utility with three modifications, summarized in the value function below.[2]

3.1 Reference dependence

Outcomes are evaluated as gains or losses relative to a reference point — usually the status quo — not in absolute wealth terms. A salary cut from $100k to $90k feels worse than a raise from $80k to $90k, despite identical end states.

3.2 Loss aversion

Losses loom larger than equivalent gains. Subjects refuse 50/50 gambles to win $100 / lose $100. They typically require winning ~$200 to accept losing $100.

3.3 Diminishing sensitivity

The difference between $0 and $100 feels larger than between $1,000 and $1,100. Same on the loss side. The function flattens away from the reference point.

3.4 Probability weighting

People overweight small probabilities (lottery tickets, plane crashes) and underweight large ones. The π function in PT replaces objective probability with subjective decision weights.

4 · The Catalogue of Biases

Below, the most replicated and operationally relevant deviations.

5 · Heuristics

Heuristics are mental shortcuts. They are not always wrong — Gigerenzer (2008) argues they are often ecologically rational — but they fail in predictable conditions.

5.1 Availability

Asked whether more English words start with "k" or have "k" as third letter, most say the former. The reverse is true; words starting with "k" are simply easier to recall.

5.2 Representativeness

Linda is bright, single, deeply concerned with social justice. Asked which is more probable: (A) Linda is a bank teller, (B) Linda is a bank teller and active in the feminist movement — most subjects pick B. But P(A∩B) ≤ P(A) by definition.

5.3 Anchoring & adjustment

Real-estate agents shown the same house with two different list prices give appraisals that vary by 10–15% in the direction of the anchor — even when they deny being influenced.

6 · Choice Architecture

Thaler & Sunstein (2008) coined "nudge" — a non-coercive change to the choice environment that predictably alters behavior, while preserving freedom of choice.[3]

6.1 Defaults

What happens if the user does nothing? In 401(k) enrollment, opt-in regimes typically yield ~40% participation; opt-out regimes typically yield ~90%. Same employees, same plan, different default — 50 percentage point difference.

6.2 Salience

Calorie labels on menus. Fuel-economy labels in MPG vs. gallons-per-mile. The choice doesn't change; the cognitive cost of choosing well does.

6.3 Friction

Adding a single click reduces uptake meaningfully. Removing one increases it. Amazon's 1-click patent (1999) was, in behavioral terms, a friction-removal nudge.

7 · Case · Save More Tomorrow

Thaler & Benartzi (2004) designed a program in which employees pre-commit to direct future raises into 401(k) contributions.[4] By tying the increase to a pay raise, the worker never experiences nominal income loss.

Field implementation: 78% of offered employees enrolled. Average savings rate climbed from 3.5% to 13.6% over 40 months. The program leveraged three behavioral levers simultaneously: present bias (commitment device), loss aversion (framed as gain not loss), and status quo (auto-escalation).

Illustrative placeholder behavioral economics lab image (picsum.photos)

8 · Case · Organ Donation Defaults

Johnson & Goldstein (2003) compared organ donor registration rates across European countries.[5]

Cultural attitudes toward death don't differ this much between Germany and Austria. The form does. The lesson: the default is a policy choice. Pretending it isn't one is itself a choice.

9 · Critiques

9.1 The replication crisis

Many classic findings (ego depletion, social priming, power posing) have failed to replicate at full effect size. Behavioral economics has been more robust than social psychology overall, but some textbook results — including some priming effects in Thinking, Fast and Slow — should be held loosely. Kahneman acknowledged this publicly in 2017.

9.2 The WEIRD sample problem

Henrich et al. (2010): most behavioral findings come from Western, Educated, Industrialized, Rich, Democratic samples — and often from undergraduates. Generalizing to the global population requires care. Cross-cultural replications find both robust effects (loss aversion) and culturally bounded ones (some framing effects).

9.3 Are nudges enough?

Critics (Loewenstein, Chater) argue nudges treat symptoms while leaving structural problems untouched. A default opt-in to a bad pension plan is still a bad pension plan. Sludge — the inverse of nudge, friction added to bad choices — is sometimes the better lever.

10 · Recommended Reading & Watching

Books

Kahneman — Thinking, Fast and Slow (2011). The canon.
Thaler & Sunstein — Nudge (2008). The policy bible.
Thaler — Misbehaving (2015). The history of the field.
Ariely — Predictably Irrational (2008).
Cialdini — Influence (1984). Persuasion's behavioral roots.
Sutherland — Alchemy (2019). Behavioral econ for marketers.
Heath & Heath — Switch (2010). Behavior change in practice.

YouTube

Daniel Kahneman — "Talks at Google"
Richard Thaler — Nobel Lecture, 2017
Stanford GSB — behavioral economics seminars
Khan Academy — behavioral econ playlist
Dan Ariely — TED talks