Questions: Reference-Dependent Preferences

Question 1

In the Koszegi-Rabin model, the reference point is determined by...

• A. An arbitrary anchor provided by the experimenter
• B. The person's rational expectations about outcomes, formed before the decision
• C. The highest outcome the person has ever experienced
• D. A fixed biological set point

Answer: (B)

Koszegi and Rabin's key innovation was to endogenize the reference point as rational expectations. Rather than assuming the reference point is the status quo or some arbitrary value, their model proposes that people form expectations about what they will receive (based on available information) and then evaluate actual outcomes relative to those expectations. This makes the reference point endogenous to the economic environment — it responds to information, prices, and probabilities — while preserving the gain-loss evaluation framework from prospect theory.

Question 2

Reference-dependent preferences predict that consumers will be more price-sensitive when a price increase exceeds their expectations than when the same high price has always been the norm.

• T. True
• F. False

Answer: True

If the reference point is expectations, a consumer who expects to pay $5 and encounters a $7 price experiences a $2 loss relative to the reference point — activating loss aversion and reducing demand more than the price change alone would predict. A consumer who always expected $7 has no loss relative to expectations and responds based on standard price sensitivity alone. This asymmetry between expected and unexpected price changes is a distinctive prediction of reference-dependent models and has been confirmed in field data.

Question 3

How does reference-dependence explain the finding that New York City taxi drivers work fewer hours on high-wage days?

Model answer: Camerer et al. found that taxi drivers set a daily income target (reference point) and stop working once they reach it. On high-fare days (rainy, busy), they reach the target quickly and quit early; on low-fare days, they work longer to approach the target. This is the opposite of the standard labor supply prediction (work more when wages are high) but consistent with reference-dependent preferences: once income exceeds the reference point, additional earnings provide diminishing gain-loss utility, reducing the incentive to continue.

This finding has been debated — Farber argues that measurement issues and selection effects complicate the original analysis — but the theoretical point stands: reference-dependent preferences can produce 'target earning' behavior where labor supply slopes negatively with the daily wage. The Koszegi-Rabin model would predict this if the reference point is a daily income expectation, because surpassing the expectation provides diminishing marginal gain-loss utility while falling short of it produces steep marginal loss.