Money illusion

Money illusion is the tendency to judge money by the number printed on it rather than by the goods or services that number can buy. The first number is nominal value. The second is real value, which adjusts for a relevant change in prices.

The bias does not require someone to be unaware of inflation. A person can understand that prices have changed and still let the more visible currency figure steer the comparison.

When a higher price is lower in real terms

This studio scene is an illustrative hypothetical. It is neither a documented case nor a research procedure.

A community pottery studio paid $3,000 for a year of kiln maintenance. Its renewal quote is $3,120. Several board members see the 4% increase and call the vendor more expensive.

The studio's bookkeeper adds another number. A price index chosen for comparable equipment services rose 6% over the same period. Expressed in last year's dollars, the renewal costs about $2,943:

$3,120 ÷ 1.06 = $2,943.40

On that benchmark, the contract's real price fell by about 1.9%, even though its nominal price rose. This does not prove that the renewal is a good deal. The index may fit imperfectly, and coverage, repair times, taxes and vendor quality still matter. It does show why the sticker change cannot answer the question by itself.

What counts as money illusion

An economic comparison needs three pieces: a nominal amount, a price-level benchmark for the same period and a real calculation. A larger currency amount may represent a gain, no change or a loss in purchasing power. The direction depends on how fast the relevant prices moved.

This is why money illusion is clearest when two outcomes have the same purchasing power but are presented with different nominal figures, or when a nominal change and its inflation-adjusted change point in different directions. Merely noticing that a wage, balance or price changed is not the bias.

In their foundational paper, Eldar Shafir, Peter Diamond and Amos Tversky used income, transaction, consumption and contract problems to test nominal-versus-real judgments. Their account allowed people to use both representations. The error could emerge from how those views competed in a particular decision, rather than from total ignorance of real value (Shafir, Diamond, and Tversky, 1997).

A later online study with 604 participants replicated the four focal problems. Its extension did not find support for the idea that knowing or correctly estimating inflation was associated with less money illusion (Ziano et al., 2021). These were structured judgment tasks, so they establish neither a population rate nor a universal effect on financial behavior.

What experiments reveal, and what they leave unsettled

Price-setting under nominal and real displays

Fehr and Tyran tested money illusion in an incentivized price-setting game. Participants faced a fully anticipated monetary shock and made decisions with payoffs shown in nominal or real terms. After a negative shock, prices and expectations adjusted more slowly under the nominal display. The difference after a positive shock was much smaller. Because each participant's best response depended on other participants' choices, a small individual pull toward nominal figures could be magnified by coordination (Fehr and Tyran, 2001).

The explanation remains contested. Petersen and Winn found no evidence that participants sacrificed real payoff to choose a higher nominal payoff. They argued that many participants used nominal figures as a shortcut among choices that already maximized real payoff, while cognitive difficulty and the game's best-response structure explained much of the delay (Petersen and Winn, 2014). Fehr and Tyran replied that both datasets still showed nominal effects on expectations and the same negative-versus-positive shock asymmetry (Fehr and Tyran, 2014).

That exchange supports a conditional mechanism. A nominal display can simplify a difficult evaluation and shape expectations, especially when decisions are interdependent. It does not support calling the effect cognitive laziness, or treating a face-value shortcut as the only cause of slow adjustment.

A separate search-economy experiment also shows why the boundary matters. Money-stock increases produced broadly proportional price adjustment without real effects in that setting, while decreases did not adjust in the same way (Duffy and Puzzello, 2014). One laboratory result cannot stand in for every market or monetary change.

A neural signal without a behavioral difference

Weber and colleagues placed 24 participants in low- and high-nominal conditions that preserved the same purchasing power. Activity in the ventromedial prefrontal cortex was greater for the higher nominal rewards. Yet task performance and real earnings did not differ between the conditions (Weber et al., 2009). The study is evidence that nominal representation can enter valuation-related neural responses. It is not proof that a brain signal causes broad consumer or market behavior.

Where the idea travels carefully

Wages

A nominal raise can be a real pay cut when prices rise faster than pay. The opposite issue appears with explicit wage cuts: people may resist a smaller number even when another inflation path would have produced the same or worse purchasing power.

Money illusion is only one possible explanation. Kahneman, Knetsch and Thaler's survey research found strong fairness asymmetries around wage and price changes, including a difference between explicit nominal wage cuts and real adjustment through inflation (Kahneman, Knetsch, and Thaler, 1986). Those judgments can reflect losses, reference points and perceived employer intent.

An incentivized gift-exchange experiment sharpened that alternative. Under inflation, an intentional nominal wage increase affected worker effort when a human employer chose it, but the effect was absent when a computer set the wage. The authors attributed the response to reciprocity and inferred intention rather than a simple failure to understand real pay (Grundmann, Giamattei, and Lambsdorff, 2019). A wage response should therefore be explained, not labeled on sight.

Transactions, contracts and asset prices

The Shafir problems and their 2021 replication support nominal-real differences in consumer transactions and contract judgments. Because the choices were hypothetical, they do not quantify aggregate spending or show that every contracting party makes the same error.

Asset-market research offers a different kind of evidence. Brunnermeier and Julliard modeled homebuyers who compare rents with nominal mortgage payments without fully allowing for inflation's effect on future real payments. Their time-series results were consistent with that mechanism contributing to housing mispricing (Brunnermeier and Julliard, 2008). Cohen, Polk and Vuolteenaho found stock-market patterns consistent with investors discounting real cash flows at nominal rates (Cohen, Polk, and Vuolteenaho, 2005).

Both studies depend on models and observational market data. They support carefully bounded applications of money illusion, not a claim that every investor performs the same mistaken calculation.

Common misunderstandings

“A higher number means money illusion.” No. If income rises faster than the relevant price level, real income rises too. The mismatch matters only after both changes are compared.

“Knowing the inflation rate removes the problem.” Not reliably. Inflation knowledge is necessary for some real calculations, but the replication extension found no simple relationship between correct inflation estimates and reduced money illusion.

“All sticky prices or wage disputes come from nominal thinking.” Strategic expectations, calculation difficulty, fairness, losses and inferred intentions can produce similar behavior. The cause has to be tested in the relevant setting.