Gift Giving

1
Gift Giving
2
Your last gift.

• What was the last gift you received (money
  counts)?
• Who gave it to you (parent, grandparent, friend)?
  
• What would you estimate the price the person paid
  to buy it?
• What is the amount of cash such that you are
  indifferent between the gift and the cash, not
  counting the sentimental value of the gift?
• Why do you value or not value this gift?

3
Summary of gift survey.

• If you are an aunt, dont give socks!
• Beware of taking a girlfriend on holiday.
• Cash from parents/grandparents was appreciated.
• A bit surprised that a few that a bad DVD cost
  them 5, but they wanted 10 for it.
• Some thought that the gifts they gave increased
  value.
• Others thought they would decrease value but the
  gave them anyway. Why?
• Some listed a bad gift as one that had little
  value or showed little thought even though the
  yield was okay. Why?
• Any suggests for improvement of the survey (such
  as understanding email me).

4
Todd Kaplan and Bradley Ruffle (2008) "In search
of welfare-improving gifts".

• Motivation
• claim that gift giving is welfare reducing rests
  on several assumptions
• the giver does not perfectly know the recipients
  preferences.
• gifts cannot be costlessly refunded.
• gift recipients possess full information as to
  whereabouts of goods they desire
• gift recipients are able to obtain such goods
  costlessly
• Kaplan and Ruffle (KR) break with this literature
  by relaxing assumptions 3 4
• 1) they add uncertainty about the existence
  location of goods and
• 2) search costs to resolve this uncertainty
• importance of search-cost savings in modern gift
  giving can be heard in common expressions of
  gratitude upon receipt of a gift "where did you
  find it? I've looked all over for this item."

5
Simplified Model

• There is a giver and a receiver.
• The giver is at a store and has to decide whether
  or not to buy a gift for the receiver.
• The receiver would have to spend c to visit the
  store.
• The gift costs p to purchase.
• There is an a chance of the good having value v
  (gtp) to the receiver (otherwise it is worth 0).

6
Two ways of getting the good

• Shopping the receiver travels to the store and
  buys the good, the social benefit is a (v-p)-c
• Gift Giving the giver gives the good to the
  receiver, the social benefit is a v-p
• When is gift giving better than shopping?
• a v-pgt a (v-p)-c
• Or cgt(1- a )p
• Thus, we have gift giving if cgt(1- a )p and a vgtp

Gift giving is better than shopping
Giving is not waste of money
7
Interpretation of requirements

• Gifts when cgt(1- a )p and a vgtp
• Grandmother effect when a is low, give cash
  since a vltp.
• When a is high, gifts are better option than
  buying it oneself best friends.
• When c is high, gifts are better.
• v doesnt affect which method is superior.
• Examples what is the social value of gift
  giving, a v-p, and shopping, a (v-p)-c, when
• (c,v,p,a)(1,2,1,.6), (1,3,1,.6),(1,6,2,.3),(1,8,2
  ,.3)
• gggt0gtshop, gggtshopgt0, shopgt0gtgg, shopgtgggt0

8
Why is gift giving still made?

• If someone has better information, lower search
  costs or a location advantage, why do they need
  to give gifts?
• Cant they just charge money?

9
Why not trade instead of give?

• Cant the giver simply make a profit buying from
  the store and selling to the receiver?
• In such a case, the receiver would only buy the
  good if it is worth v (with probability a).
• The receiver would bargain to purchase the good
  for a price less than v (buying at v would leave
  him indifferent).
• Go back to (c,v,p,a)(1,2,1,.6).
• If the giver spends 1, and sales it to the buyer
  for 1.9 (ltv2), he would on average receive 1.14
  for a profit of .14.
• How much must the giver get from the receiver in
  order to make a profit?
• This lack of trade is called the hold-up problem.

10
Why not trade (part 2)?

• We can interpret our model as an information
  acquisition model.
• The giver knows more than the about the good.
• The giver knows this is something the receiver
  potential wants (with prob a ).
• The giver may at other times see other products
  with lower a .
• The cost c is what it costs for the receiver to
  learn whether it is something he wants.
• Trade would not solve this basic problem, since
  the receiver would still have to spend c and
  without doing so the giver would have incentive
  to push unwanted products. (The
  stereo/car/fashion salesman.)

11
Experiment game

• We ran an experiment on what is called the
  Beer-Quiche Game (Cho Kreps, 1987).
• Proposer has 2/3 chance of being strong.
• He can eat Beer or Quiche.
• Strong types like Beer. Weak types like Quiche.
• Responder can fight or flee. Responders dont
  want to fight a strong type.

12
Signalling in the LabTreatment 1
Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder
Flee Fight
Beer (Strong) 2.00, 1.25 1.20, 0.75
Quiche (Strong) 1.00, 1.25 0.20, 0.75
Beer (Weak) 1.00, 0.75 0.20, 1.25
Quiche (Weak) 2.00, 0.75 1.20, 1.25

• For a strong proposer, (Beer, flee)gt(Beer,
  fight)gt(Quiche, flee)gt(Quiche, fight).
• For a weak proposer, (Quiche, flee)gt(Quiche,
  fight)gt(Beer, flee)gt(Beer, fight).
• Strong chooses Beer and Weak chooses Quiche

13
Signalling in the LabTreatment 1
Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder
Flee Fight
Beer (Strong) 2.00, 1.25 1.20, 0.75
Quiche (Strong) 1.00, 1.25 0.20, 0.75
Beer (Weak) 1.00, 0.75 0.20, 1.25
Quiche (Weak) 2.00, 0.75 1.20, 1.25

• Responder now knows that Beer is the choice of
  the strong type and Quiche is the choice of the
  weak type.
• For Beer he flees, for Quiche he fights.

14
Signalling in the LabTreatment 1
Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder
Flee Fight
Beer (Strong) 2.00, 1.25 1.20, 0.75
Quiche (Strong) 1.00, 1.25 0.20, 0.75
Beer (Weak) 1.00, 0.75 0.20, 1.25
Quiche (Weak) 2.00, 0.75 1.20, 1.25

• So the equilibrium is
• For strong, (Beer, Flee)
• For weak, (Quiche, Fight)
• This is called a separating equilibrium.
• Any incentive to deviate?

15
Signalling in the LabTreatment 1
Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder
Flee Fight
Beer (Strong) 2.00, 1.25 1.20, 0.75
Quiche (Strong) 1.00, 1.25 0.20, 0.75
Beer (Weak) 1.00, 0.75 0.20, 1.25
Quiche (Weak) 2.00, 0.75 1.20, 1.25
32
13
What did you do? In the last 5 rounds, there were
32 Strong and 13 Weak proposers
16
Treatment 2.
Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder
Flee Fight
Beer (Strong) 1.40, 1.25 0.60, 0.75
Quiche (Strong) 1.00, 1.25 0.20, 0.75
Beer (Weak) 1.00, 0.75 0.20, 1.25
Quiche (Weak) 1.40, 0.75 0.60, 1.25

• Can we have a separating equilibrium here?.
• If the proposer is weak, he can choose Beer and
  get 1.00 instead of 0.60.

17
Treatment 2.
Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder
Flee Fight
Beer (Strong) 1.40, 1.25 0.60, 0.75
Quiche (Strong) 1.00, 1.25 0.20, 0.75
Beer (Weak) 1.00, 0.75 0.20, 1.25
Quiche (Weak) 1.40, 0.75 0.60, 1.25

• Can choosing Beer independent of being strong or
  weak be an equilibrium?
• Yes! The responder knows there is a 2/3 chance of
  being strong, thus flees.
• This is called a pooling equilibrium.

18
Treatment 2.
Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder
Flee Fight
Beer (Strong) 1.40, 1.25 0.60, 0.75
Quiche (Strong) 1.00, 1.25 0.20, 0.75
Beer (Weak) 1.00, 0.75 0.20, 1.25
Quiche (Weak) 1.40, 0.75 0.60, 1.25
4
30
3
8

• Did we have a pooling equilibrium?
• In the last 5 rounds there were 34 strong
  proposers and 11 weak proposers.
• Do you think there is somewhat to help the
  pooling equilibrium to form?

19
Treatment 2.
Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder
Flee Fight
Beer (Strong) 1.40, 1.25 0.60, 0.75
Quiche (Strong) 1.00, 1.25 0.20, 0.75
Beer (Weak) 1.00, 0.75 0.20, 1.25
Quiche (Weak) 1.40, 0.75 0.60, 1.25
23
14
3

• At Texas AM, the aggregate numbers were shown.
• In the last 5 periods, 23 proposers were strong
  and 17 weak.

20
Signalling game

• Spence got the Nobel prize in 2001 for this.
• There are two players A and B. Player A is
  either strong or weak.
• Player B will chose one action (flee) if he knows
  player A is strong
• and another action (fight) if he knows player A
  is weak.
• Player A can send a costly signal to Player B (in
  this case it was to drink beer).

21
Signal

• For signalling to have meaning,
• we must have either cost of the signal higher for
  the weak type.
• Or the gain from the action higher for the strong
  type.

22
Types of equilibria

• Separating.
• Strong signal
• Weak dont signal.
• Pooling.
• Strong and weak both send the signal.
• (or Strong and weak both dont send the signal.)

23
Types of equilibria

• Player A is the proposer and B the responder.
• The types of player A are s and w.
• Let us normalize the value to A when B fights to
  0.
• The values to A when B flees are Vs and Vw.
• The cost to signalling (drinking beer) are Cs and
  Cw.
• We get a separating equilibria if Vs-Csgt0 and
  Vw-Cwlt0.
• We get a pooling equilibria if Vs-Cslt0 and
  Vw-Cwlt0 (no one signals).
• We may also get a pooling equilibria if Vs-Csgt0
  and Vw-Cwgt0 and there are enough s types.
• For this to happen, there must be enough s types
  such that the expected payoff of B is higher
  fleeing than fighting.

24
Treatment 2 Other pooling?.
Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder Payoffs Proposer, Responder
Flee Fight
Beer (Strong) 1.40, 1.25 0.60, 0.75
Quiche (Strong) 1.00, 1.25 0.20, 0.75
Beer (Weak) 1.00, 0.75 0.20, 1.25
Quiche (Weak) 1.40, 0.75 0.60, 1.25

• How about both proposers eat quiche and the
  responder flees? Is this an equilibrium?
• If responders think anyone who drinks Beer must
  be weak.
• Cho-Kreps introduce an intuitive criteria that
  says this does not make sense.
• Any proposer drinking Beer must be strong,
  because the weak type can only lose from doing so.

25
How does this relate to gift giving?

• Gift giving can be wasteful. (Why not give ?)
• Basically, you get someone a gift to signal your
  intent.
• American Indian tribes, a ceremony to initiate
  relations with another tribe included the burning
  of the tribes most valuable possession.

26
Courtship gifts.

• Dating Advice.
• Advice 1 never take such advice from an
  economist.
• Advice 2.
• Say that there is someone that is a perfect match
  for you. You know this, they just havent figured
  it out yet.
• Offer to take them to a really expensive place.
• It would only make sense for you to do this, if
  you knew that you would get a relationship out of
  it.
• That person should then agree to go.

27
Valentines Day

• Who bought a card, chocolate, etc?
• We are forced to spend in order to signal that we
  really care.
• Say that you are either serious or not serious
  about your relationship.
• If your partner knew you were not serious, he or
  she would break up with you.
• A card is pretty inexpensive, so both types buy
  it to keep the relationship going.
• Your partner keeps the relationship since there
  is a real chance you are serious.
• No real information is gained, but if you didnt
  buy the card, your partner would assume that you
  are not serious and break up with you.

28
Higher value and/or Lower Cost

• Higher value
• You buy someone a gift to signal that you care.
• Sending a costly signal means that they mean a
  lot to you.
• For someone that doesnt mean so much, you
  wouldnt buy them such a gift.
• Lower cost
• The person knows you well.
• Shopping for you costs them less.
• They signal that they know you well.

29
Other types of signalling in the world

• University Education.
• Showing up to class.
• Praying. Mobile phone for Orthodox Jews
• Poker Raising stakes (partial).
• Peacock tails.
• Limit pricing.

30
Homework

• 2 Questions are due on Friday, Oct 24th.
• Links the them and how to submit them are from
  the webpage.