Schools of economic thought Lesson 9 of 9
Economics · Schools of economic thought
Behavioral economics: prospect theory, nudges and the evidence
How behavioral economics departs from the rational model: bounded rationality, prospect theory, mental accounting, present bias, nudges, and what replication found.
On this page
- The rational benchmark and the departure
- Prospect theory and the value function
- Present bias: a worked example
- Nudges and automatic enrolment
- What the evidence says against
- Behavioral economics compared with the standard model
- How Finabase does it
- Frequently asked questions
- Summary
- Check your understanding
- Exercise
Behavioral economics studies how people actually make economic decisions, using evidence from psychology and experiments, where standard economics assumes they optimize. It does not claim people are foolish. It claims they use shortcuts, care about gains and losses relative to a reference point, keep money in mental compartments and put off what is good for them, in ways that are predictable enough to model.
The field has three Nobel prizes behind it: Herbert Simon (1978) for research on how decisions are made inside organizations, Daniel Kahneman (2002) for bringing psychological research on judgment under uncertainty into economics, and Richard Thaler (2017) for contributions to behavioral economics. It has also changed policy, most visibly through automatic enrolment in retirement plans.
This lesson sets out the main ideas in their own terms, computes the prospect-theory value function from its published parameters, and then turns to what the field's critics and its own replication studies have found. Investor-specific biases have their own lessons, listed below.
The rational benchmark and the departure
Standard economics models a decision maker who has stable preferences, weighs every option by its expected payoff, uses all available information, discounts the future at a constant rate and treats a dollar as a dollar wherever it comes from. That model is a powerful benchmark: it predicts that people respond to prices and incentives, which they do.
Behavioral economics keeps the benchmark and documents systematic departures from it. Its main building blocks:
- Bounded rationality. Herbert Simon argued in 1955 that people have limited information and computing power, so they satisfice: they search until they find an option that meets an aspiration level, then stop.
- Heuristics and biases. Kahneman and Amos Tversky showed that mental shortcuts, such as judging likelihood by how easily examples come to mind, produce predictable errors.
- Prospect theory. Choices under risk depend on gains and losses, not final wealth.
- Mental accounting. Thaler (1985) showed that people sort money into accounts with different rules, such as a tax refund that is spent more freely than salary.
- Present bias. People weigh "now" much more heavily than any future date, so plans to save or diet tomorrow are dropped when tomorrow arrives. David Laibson's 1997 model of this "hyperbolic" discounting explained why people under-save yet buy illiquid assets to commit themselves.
- Social preferences. People care about fairness and reciprocity, not only their own payoff.
Prospect theory and the value function
In 1979 Kahneman and Tversky proposed that people evaluate a risky choice with a value function over gains and losses from a reference point, usually the status quo. It has three features: it is concave for gains (each extra dollar of gain adds less), convex for losses (each extra dollar of loss hurts less) and steeper for losses than for gains. In 1992 Tversky and Kahneman estimated its parameters from their experiments.
Two more predictions follow. Because the curve is convex for losses, people tend to gamble to avoid a sure loss, the root of the The disposition effect: selling winners and holding losers in investing. And because the reference point matters, the same outcome can be framed as a gain or a loss and chosen differently (The framing effect and narrow framing: same facts, different choices). The lessons on Loss aversion: why losses hurt twice as much as gains please and Mental accounting: how labels on money change what investors do go deeper.
Present bias: a worked example
Nudges and automatic enrolment
Thaler and Cass Sunstein's Nudge (2008) argued that since the way choices are presented (choice architecture) inevitably affects them, designers should pick presentations that help people, while keeping every option open. A nudge changes behavior without banning options or changing financial incentives much: a default, a reminder, a simpler form.
The best-known evidence is from retirement saving. Madrian and Dennis Shea studied a large US firm that switched new hires from opting in to its 401(k) plan to being enrolled automatically at a 3% contribution invested in a money market fund, with the right to opt out.
Participation more than doubled. But the same inertia worked the other way: 76% of auto-enrolled participants stayed at the 3% default rate, and 61% of the auto-enrolled cohort kept both the default rate and the default fund, while three-quarters of participants who had enrolled themselves contributed 6% or more. A default can raise participation and lower the amount saved by people who would otherwise have chosen more, so the level of the default matters as much as its existence.
What the evidence says against
Behavioral economics has strong evidence behind its core findings and a real problem with some of its famous ones.
- Replication. The Open Science Collaboration (2015) repeated 100 psychology studies: 97% of the originals had significant results, but only 36% of the replications did, and effects were half as large. A 2016 project replicating 18 laboratory experiments published in two top economics journals found a significant effect in the same direction in 11 (61%), with effects averaging 66% of the originals. Several popular findings used to illustrate behavioral ideas have not held up.
- Publication bias in nudging. A 2022 meta-analysis of more than 200 studies found an average nudge effect of Cohen's d = 0.45 and a moderate bias toward publishing positive results; a reanalysis the same year argued that once that bias is corrected, the evidence for an average nudge effect disappears.
- Scale. Stefano DellaVigna and Elizabeth Linos (2022) compared nudges published in journals with every trial run by two US government nudge units. The journals' average effect on take-up was 8.7 percentage points (a 33.5% increase); the nudge units' was 1.4 points (8.1%), still significant but much smaller.
- Markets discipline some biases. Critics in the rational tradition argue that experienced traders, competition and arbitrage weed out costly mistakes, so lab biases may not move market prices.
- Too many parameters. With a bias for every occasion, critics say, behavioral models can explain anything after the fact unless they commit to predictions in advance.
The fair summary: loss aversion, present bias, the power of defaults and inertia are robust and matter for policy; many individual priming and framing results are fragile, and typical effects in the field are smaller than the first papers suggested.
Behavioral economics compared with the standard model
| Topic | Standard model | Behavioral finding |
|---|---|---|
| Risk | Expected utility of wealth | Gains/losses; losses weigh ~2.25× |
| Time | Constant discounting | Present bias; plans reverse |
| Money | Fungible | Mental accounts by source and use |
| Defaults | Irrelevant if opt-out free | 37% → 86% participation (one firm) |
| Information | Used fully | Limited attention; satisficing |
How Finabase does it
Finabase does not try to nudge anyone into a trade, and it has no behavioral scoring of your decisions. What it offers are tools that work against some of these biases if you choose to use them: a portfolio that shows every holding's gain or loss against its cost, which makes the reference point explicit; paper trading, for practicing decisions without money at stake; and price alerts, which let a plan made in advance trigger the next look instead of the screen. The Cognitive biases in investing: what the research actually shows lesson and the investor-psychology lessons, such as Overconfidence and overtrading: the research on trading too much and The endowment effect and status quo bias: why you keep what you own, apply these ideas to investing.
Frequently asked questions
What is behavioral economics in simple terms?
It is economics that uses evidence on how people really decide, including shortcuts, fear of losses and procrastination, instead of assuming everyone optimizes. Its models aim to predict these departures, not just list them.
What is the difference between behavioral economics and behavioral finance?
Behavioral finance is the part of behavioral economics that studies investors and markets, such as why investors hold losers too long or chase recent winners. Behavioral economics also covers saving, consumption, labor, health and public policy.
What is a nudge?
A nudge is a change in how a choice is presented, such as a default option or a reminder, that steers behavior without forbidding options or changing incentives much. Automatic enrolment in a retirement plan with the right to opt out is the standard example.
Has behavioral economics been debunked?
No, but parts of it have been weakened. Core findings such as loss aversion, present bias and the power of defaults have held up, while a number of famous lab results failed to replicate, and nudge effects in large government trials are much smaller than in published papers.
Summary
Behavioral economics replaces the fully rational decision maker with one who satisfices, judges outcomes as gains and losses, keeps mental accounts and favors the present. Prospect theory's value function, with Tversky and Kahneman's 1992 parameters, makes losses weigh about 2.25 times as much as gains, so a 50/50 bet needs a $251 upside to offset a $100 downside. The field's biggest policy success, automatic enrolment, more than doubled participation at the firm Madrian and Shea studied but left most savers at a low default. Its critics, and its own replication projects, show that many early findings were overstated and that nudges at scale are real but small.
Check your understanding
4 questions on this lesson. Each answer comes with the reason, and nothing is scored.
Using the Tversky–Kahneman 1992 parameters, a person is offered a 50/50 bet: win $150 or lose $100. What does prospect theory predict?
- They accept, because the expected value is +$25
- They refuse, because a $100 loss feels like about −129 units and a $150 gain like only about +82
- They are indifferent
Show the answer
B. They refuse, because a $100 loss feels like about −129 units and a $150 gain like only about +82 150^0.88 ≈ 82 while 2.25 × 100^0.88 ≈ 129. With losses weighted 2.25 times, the gain would need to be about $251 to balance a $100 loss, so a bet with positive expected value is turned down.
A firm switches new hires to automatic 401(k) enrolment at a 3% contribution. Before, many self-enrolled workers chose 6% or more. What does the Madrian–Shea evidence suggest will happen?
- Participation rises sharply, but many people stay at 3%, so some save less than they would have
- Participation is unchanged, because people already knew they could enrol
- Everyone raises their contribution to 6% within a year
Show the answer
A. Participation rises sharply, but many people stay at 3%, so some save less than they would have At the firm they studied, participation among new hires rose from 37% to 86%, but 76% of auto-enrolled participants stayed at the 3% default. Inertia raises participation and also anchors people to the default rate.
Someone plans in January to start saving 10% of pay "from next month", and every month postpones it again. Which idea explains this best?
- Loss aversion
- Present bias: the immediate cost is weighted much more heavily when it is now than when it was planned
- Mental accounting
Show the answer
B. Present bias: the immediate cost is weighted much more heavily when it is now than when it was planned With present bias, the trade-off looks worth it when both cost and benefit are in the future, but when "next month" becomes "now" the cost gets extra weight and the plan reverses. This is the procrastination pattern Laibson’s model describes.
A journal article reports that a reminder letter raised take-up of a benefit by 9 percentage points. A government nudge unit plans to send the same kind of letter nationwide. What does the evidence suggest it should expect?
- About 9 points, since the study was randomized
- A smaller effect: published nudges averaged 8.7 points, but nudge-unit trials averaged 1.4
- No effect at all, since nudges do not replicate
Show the answer
B. A smaller effect: published nudges averaged 8.7 points, but nudge-unit trials averaged 1.4 DellaVigna and Linos found that nudge-unit trials, which include every result rather than the published ones, had effects about one-sixth the size of those in journals: still positive and significant, but small.
Exercise
One task to do in Finabase. Nothing is graded or saved.
Open paper trading and buy a small paper position in a stock or fund of your choice. Before you leave, write down two prices, one 10% below your entry and one 10% above, and what you will do at each. Set a price alert at both. When an alert fires, compare what you feel like doing with what you planned. Did the loss feel bigger than the equal gain, and did you want to wait for the price to come back?
Open paper tradingWhat you should find
Loss aversion predicts the paper loss feels about twice as strong as the equal gain (λ = 2.25 in the 1992 estimates), and the urge to wait for a loser to recover is the disposition effect. A plan written in advance is a commitment device against present bias: when the plan and the urge differ, the urge is usually the bias.
Sources and further reading
- Simon, A behavioral model of rational choice, Quarterly Journal of Economics 69(1), 1955
- Kahneman and Tversky, Prospect theory: an analysis of decision under risk, Econometrica 47(2), 1979
- Tversky and Kahneman, Advances in prospect theory: cumulative representation of uncertainty, Journal of Risk and Uncertainty 5(4), 1992
- Thaler, Mental accounting and consumer choice, Marketing Science 4(3), 1985
- Laibson, Golden eggs and hyperbolic discounting, Quarterly Journal of Economics 112(2), 1997
- Thaler and Sunstein, Nudge: Improving Decisions about Health, Wealth, and Happiness (Yale University Press, 2008)
- Madrian and Shea, The power of suggestion: inertia in 401(k) participation and savings behavior, NBER Working Paper 7682 (2000); Quarterly Journal of Economics 116(4), 2001, retrieved 2026-11-18
- DellaVigna and Linos, RCTs to scale: comprehensive evidence from two nudge units, Econometrica 90(1), 2022 (NBER Working Paper 27594), retrieved 2026-11-18
- Open Science Collaboration, Estimating the reproducibility of psychological science, Science 349(6251), 2015, retrieved 2026-11-18
- Camerer et al., Evaluating replicability of laboratory experiments in economics, Science 351(6280), 2016, retrieved 2026-11-18
- Mertens et al., The effectiveness of nudging: a meta-analysis of choice architecture interventions, PNAS 119(1), 2022, retrieved 2026-11-18
- Maier et al., No evidence for nudging after adjusting for publication bias, PNAS 119(31), 2022, retrieved 2026-11-18
- NobelPrize.org — Prize in Economic Sciences 1978 (Simon), 2002 (Kahneman, Smith) and 2017 (Thaler), retrieved 2026-11-18
