Predictably Misbehaving

Lecture 2: Reference-dependent Preferences

Joshua Foster at UW-Oshkosh

The Size-Contrast Illusion

The Ponzo Illusion

Comparative Perceptions

Our brains make relative comparisons, not absolute.

Brightness, loudness, temperature...all relative comparisons.
So are many of our experiences.

We find comparative judgments much easier to make.

It's easy to tell which of two buckets of water is warmer.
It's hard to tell their absolute temperature.

Today's Major Takeaway

Evaluations of economic outcomes are heavily influenced by comparisons.

One gets strong feelings comparing incomes.

Say a friend makes $1000 more than you.

It's hard to judge how much $1000 will help.

If absolute consumption is what made people happy...

We should be happier than people 100 years ago.
(e.g. electricity, indoor plumbing, entertainment)

Traditional Economics

Traditional Utility

Traditional utility depends on absolute valuation.

The more consumption, the happier you are.
Consumption has diminishing marginal utility.

Modifications

Reference-dependent Utility

Reference-dependent (RD) utility depends on relative valuation.

RD Utility depends on what they had before.
Additionally, they really dislike losing what they had.

Loss Aversion

For the consumption of something particular, people dislike losses relative to their reference point more than they like same-sized gains.

The inclusion of loss aversion is one of the most important properties of reference-dependent utility.

Loss Aversion

"How could we test for loss aversion?"

Willingness to trade one's current position for an alternative.
Preferences over risky gambles.

Compelling studies with evidence of loss aversion:

Kahneman, Knetsch, and Thaler (1990, 1991)
And many others...

Results: once a person comes to possess something they (almost) immediately value it more than before when they did not possess it.

Evidence of Loss Aversion

Time: 32:07 - 33:55

Endowment Effect

Endowing someone with a good almost instantaneously makes them value it more highly.

Experiments consistently find a major gap in prices:

Selling prices tend to be $\sim 2\times$ the buying prices.

Testing the Endowment Effect

Experimental Procedure:

Randomly give half of the subjects mugs.
"Owners" and "non-owners" both examine them.
Elicit buying and selling prices.

The Difficulty

Getting people to honestly state their value for a mug.

The experiment must be incentive compatible.
Otherwise, prices will not reflect their true valuations.

Becker-DeGroot-Marschak (BDM) Procedure

Each non-owner receives the following table to the left.

Subjects are told a price will be selected randomly.

Cannot to influence price, thus best to tell the truth.

Endowment Effect $=$ Reference-dependence $+$ Loss Aversion

	$\Delta$ in Mugs	$\Delta$ in Money
Owners	1	0
Non-owners	0	0

Selling a mug leads to

Loss of mug (-1)
Gain of money (+$x$)

Buying a mug leads to

Loss of money (-$y$)
Gain of mug (+1)

Loss aversion predicts that $x>y$. People need extra incentives to let go of what they already have.

What does traditional economis have to say?

There are potential measurement issues:

Owners are slightly richer.
Thus an income effect could explain this result.

Additional experiments were run to find out.

Meet the "Choosers"

Each non-owner receives one of the following tables.

Ownersvs.Choosers

At each $ amount, choose...the mug, or the money.

Endowment Effect Persists

Kahneman, Knetsch, and Thaler (1991)

Owners		Non-owners		Choosers
7.12	$>$	2.87	$\approx$	3.12

Major Takeaway: Low trade volume is due to owners' reluctance to part with their mug rather than buyers' unwillingness to part with their cash.