Bestiary of Behavioral Economics/Money Illusion

Background
Irving Fisher coined the phrase in his article "Money Illusion" in the early 20th century. He defined explicitly the problem that many people think in nominal instead of real values, or as the case may be, in a mix of the two. The ease of using nominal thinking makes this problem widespread, especially when it occurs during times of hyperinflation that Irving Fisher studied. Alternative framinings of the same situation (one in nominal and one in real) can lead to different choices because people adhere to the frame presented to them. This problem leads to poor economic choices and an inadequate view of prices and costs. While inflation is the most common lense to view money illusion, sunk costs is also prominant and it comes in many forms: not accepting a lower wage, not selling at a marginal loss, investment, mental accounting, and even fairness. Though, this theory is not strictly a behavioral problem. Macroeconomists use this insight in their analysis of peoples reaction to inflation and how that effects monetary policy and compare the effects to negative interest rates. And psychological insights are very important in understanding why money illusion occurs, especially in cases where framing[] risk aversion plays a role (investing and transactions). Central to the problem of money illusion is the lack of economic education, but even with education or learning nominal values are still used.

Specific Examples
In this section, we will dissect a select couple of examples of money illusion that have been studied over the years.

Earnings
Shafir, Diamond, and Tversky, created a situation to test people's understanding of money illusion in wage earnings. There are two people who have the same college experience, same type of job, same starting wage, yet experience different percentage rates of inflation and bonus.

Person 1: $30,000 base salary, 0% inflation (the first year), 2% rise in salary (second year)

Person 2: $30,000 base salary, 4% inflation (the first year), 5% rise in salary (second year)

In strictly real terms, Person 1 is better off than Person 2 as they receive a higher wage after inflation is taken into account, while Person 2 has more "money" in a nominal sense but after taking into account inflation Person 2's money is worth less.

The questions asked of the participants, who had to evaluate Person 1 and Person 2, had to answer three questions, one on "economic terms", happiness and job attractiveness. In response to the first question about who was doing better in "economic terms", the majority sided with Person 1, which is correct in real terms. But in response to the next question on who was happier, respondents (different people from question one) said Person 2 was happier even though the majority of earlier respondents acknowledged that Person 1 was better off. This answer clearly shows the nominal bias: people assume a larger bonus means more money which in turn makes them happier. The third question reinforces this concept. Another set of respondents were asked who would be more likely to leave their job and the majority said Person 1. This response also shows nominal bias because Person 1 is actually receiving a greater bonus once inflation is taken into account.

The surveys indicated that people in fact do use nominal values to create judgements. But when a question was framed that emphasized who would be better off in "economic terms" people answered correctly saying Person 1 was better off. This means people do think in real terms when they are reminded of it, but the bias is toward nominal in all other cases.

Mental Accounting
Mental accounting[] is a topic in and of itself and more can be learned by accessing the link provided. This example with be looking at Shafir, Diamond, and Tversky's, surveys assessing people's responses to past nominal values and sunk costs. People have difficulty disassociating past costs with the present and are unsure of what to make of the connection between the two. This lead to a variety of answers and a lack of consensus. So the authors developed a second question. Two stores buy the same quantity of goods at different input prices and sell them at different output prices. One of the two buys and sells his a year later with inflation.

Store A: Bought 100 goods for $20 each and sold them at $44 each
 * -A Year Later-

Store B: Bought 100 goods for $22 each and sold them at $45 each-while experiencing 10% inflation

An overwhelming majority when asked "who made the better deal?" selected Store A over Store B. This makes sense thinking in nominal terms as you multiply the quantity times the output price and subtract the quantity times the input price. Then you repeat for both stores:

Store A:((100 quantity of goods)x($44 output price))-((100 quantity of goods)x($20 input price))= $2,400 Store B:((100 quantity of goods)x($45 output price))-((100 quantity of goods)x($22 input price))= $2,300

Under this logic (ignoring inflation) it is easy to see that the easier line of nominal thinking is very persuasive. Aknowledging inflation, the price of the input was offset by the rise inflation allowing for more money to be made by only increasing the output price by one dollar. Money illusion shows it's hand by illustrating, once again, that people instinctively think in nominal terms.

Experimental Monetary Shock
It is disputed as to whether exogenous monetary shocks in the "real world" have a real impact on output as it is still highly debated, but Fehr and Tyran ran their experiment in a lab acknowledging that field data would be filled with skepticism. The model they constructed was to have an "exogenous" and expected monetary shock. In the experiment, participants play against other players and the computer (removing any need to act off of uncertainty) to set set prices and trade with each other. They select a number (1-30) and then select the price's expectation, as well as, the confidence in their decision. To prepare to analyze the situation, they tested the people before, during and after the negative shock on the money supply. The "pre-shock" test was used for getting participants familiar with the system and using it to compare to the "post shock" price.

The results of their experiment showed:
 * With real values against computers, the shock had no almost no effect on prices with players having to make very minimal adjustments to equilibrium.
 * With nominal values against computers, the shock created a limited amount of money illusion as it became harder for participants to get back to equillibrium.
 * With real values against other participants, the shock had a limited effect on prices but there were fluxuations.
 * With nominal values against other participants, the shock had an astounding effect on prices creating such variance that the participants never reached equillibrium.

To conclude it is evident that money illusion plays a part in peoples decision making and has an even more drastic effect in a nominal frame. Even in a fully expected drop in the quantity of money still lead to large income losses for participants until they got closer to equilibrium. The same researchers conducted a study on positive economic shocks as well but the effects of money illusion were much smaller.

Conclusions
We can conclude that there is no easy solution to the problem of money illusion. Attempts to reconcile the lack in logic with economic education or to pursue policies that have very little inflation have not been very successful. Money illusion, though has immense policy and business implications especially when contracts, and even tax methods are incorrectly indexed. It is an especially complicated ordeal for businesses who wish to raise prices to combat inflation but do not want to confront the public who views in anger any nominal price rise. Macroeconomists will continue to struggle with the concept of money illusion in their explanations of short term changes in equilibrium prices. Behavioral economists still have much to research empirically about money illusion and to find evidence of it outside of the laboratory or questionaire setting.