ED 8 - IIM C Topic 2 Incentives Aug 2022
ED 8 - IIM C Topic 2 Incentives Aug 2022
Topic 2: Incentives
Theory and evidence of monetary and non-monetary incentives for pro-environmental
behaviour change
Dr Prasenjit Banerjee
University of Manchester
IIM C
Environmental Economics
Positive contribution to PG—altruism?
Direct evidence?
Well functioning
market
Homo Social
Economicus Efficiency
Market Economic
Failure Nash Incentives
Equilibrium
Conditional
Cooperation
Social preferences
Behavioural
Failure
Crowding-Out
3
Ultimatum game
A Proposer is given $10. He is asked how
much of this he wants to give to a receiver.
The receiver either accepts or rejects the
offer. If she accepts both receiver and
proposer get their share. If she rejects both
get nothing.
5
Dictator game
6
Dictator Game
• Like the ultimatum game but player 2 (Recipient) has
no choice
7
Do subjects punish unfair behaviour?
(KAHNEMAN, KNETSCH,THALER (1986).
• 2 stages
• Stage 1:
– subjects could divide a $20 pie only one of two ways,
• (1) (20 – 10, 10); or (2) (20 – 2, 2)
– The counterpart was passive => Dictator game
– Results: 76% split it equally.
• Stage 2:
– subjects were matched with two recipients
– could divide $10 (evenly) with someone who had previously
kept $10, or $12 with someone who had previously kept $18.
– Results: 74% chose the smaller pie, sacrificing $1 in payoffs to
reward the fair counterpart.
• Clearly fairness plays a role, and subjects are willing to sacrifice
small amounts of money to punish unfair behaviour.
Is anonymity a factor?
• Agents are anonymous -- no individual or social identities
– agents do not care if they are observed
• But in the experiments, subjects are not anonymous.
– subjects are fair because they are observed?
• To test this: use double-blind or double anonymous protocol
Hoffman, et al (1994, AER)
– subjects know that their actions are not observed, and they
are anonymous to each other.
• Their experiment is a dictator game
• Result: A notable increase in selfish play, with about 70% of
subjects keeping the full endowment.
Does the unearned endowment play a
role?
• Subjects received monetary endowment from the experimenter
for free
• What if they earned it, as in reality people do
• The assets in a bargain must be legitimate to produce rational
behaviour in the lab
• To test this add the following to the dictator game (Cherry et
al. 2002):
– subjects had to earn the endowment by putting real effort
(solving a GMAT quiz) – if at least 10 correct answers, they
received $40, otherwise $10
– Then decides how much to give to a recipient from the earned
endowment
Design (Cherry et al. 2002, AER)
• Subjects unfamiliar with bargaining/dictator game
• Randomly divided into 2 groups, A & B, split into 2 rooms
• 2 stages: earnings and bargaining/giving
• Bargaining: A1 (who earns $40) & A2 split into different
rooms and randomly matched with B
• A: decides a split from their earning and game ends
• 3 treatments
– Baseline: endowment given ($40 or $10) and D decides a
split
– Earnings: D earns the endowment ($40 or $10) and R knows
that D earns it (common knowledge)
– Double blind: same as ‘Earnings’ with zero visibility
Results
• Social preferences diminished with earned
wealth
• This nearly eliminated with earned wealth +
anonymity.
standard, $5
standard, $10
replication, $10 The nice side of
variation, $10 social preferences
exchange, $10 – people giving to
single blind 2, $10 others.
single blind 1, $10
double blind 2, $10
double blind 1, $10
-0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Share offered by proposers
Source: Forsythe et al. (1994), Hoffman et al. (1994), Hoffman et al (1996), Camerer (2003) 16
Reputation: a formal model
(Bénabou, R. and Tirole, J., 2006. Incentives and prosocial behavior. American economic review, 96(5), pp.1652-
1678.)
• Three motives:
extrinsic + intrinsic + reputational
– Their interaction and how incentives affect them
•
• f(v) with mean (). Its realization is private information.
Motivation and motivational crowding out
• Economists suggest that people respond to monetary incentive even if
they have other motivations (e.g. altruism, reputation)
• Evidence suggests possibility of crowding out effect
– Voluntary blood donations decreases after introducing monetary payment to
blood donation—because people unable to indulge the altruistic satisfaction
from ‘good deed’ or they have doubts about the true motives of the work
(Titmus 1970)
– A majority of Swiss residents were willing to accept repositories of nuclear
waste in their communities when no payment was offered compared to
different level of monetary payments and this willingness actually went
down when a monetary payment was offered (Frey and Oberholzer-Gee
1997)
– On the other hand, imposing fine to parents for late pick-ups of toddlers did
not change behaviour—extrinsic incentives crowd-out intrinsic motives
(Gneezy and Rustichini 2000)
Test: whether people care about
moral reputation
– People donate more when visibility increases
– Exert more effort in workplace to gain a good
reputation
– Tend to keep self-chosen savings commitment if
observed by other person
• Andreoni and Bernheim (2009)(Andreoni, J. and Bernheim,
B.D., 2009. Social image and the 50–50 norm: A theoretical and experimental
analysis of audience effects. Econometrica, 77(5), pp.1607-1636.)
– In private setting behave selfishly but in public setting
inequity averse
Altruism or 50-50 norm? Concerns for self-image?
Warm-Glow/
Crowding-Out Altruism
∂a/∂T<0 Andreoni 1995; Frey 1994;
Becker 1974; Fehr and
Titmuss 1970; Frey 1997; Frey and
Rockenbach 2004
Benabou andBowles
Jegen 2001; Tirole2008
2006
∂a/∂T>0
Participation in •at least compensate loss
pro-social action of rents
Moral •Information Rent
for reputation?
Benabou and Tirole 2006; Dana Reputation Smith and Shogren 2002; Laffont 1995; Baliga
and Maskin 2003
et al. 2003 Smith 1759; Milgrom and
Roberts 1982; Sen 1994;
Nowak & Sigmund 1998
29
Research Questions
31
Reputation: how we see it
• Perception others have about an individual’s type,
nature, or value
• People care because -- feel good ; other selfish
motive
• Here -- based on the landowners’ intrinsic valuation
for money (vT)– a high vT ,a low reputation and vice-
versa
Crowding-out
∂a/∂T <0 if RaT > vT
=> if care more about reputational gain than monetary gain
s.t. vT π(A - a) + va a + vT T a + R ≥ θ
Optimality requires
∂π(A - a)/∂a = ∂B/∂a + λ/ vT (1 + ∂R/∂a)
vTT* a* = θ – π(A - a*) – a* - R
Mechanism under Asymmetric Information about Reputation and land quality
35
EFFECT OF INCENTIVES
Effect of fine on late pick-up of pupils from
day-care centers (Gneezy and Rustichini)
• Standard economic theory predicts: attaching a fine to an activity
will reduce the amount of that activity, all else equal (the
principle of deterrence)
• This paper provides an example where that principle fails.
• A group of private Israeli day-care centers were troubled by the
frequency at which parents arrived after the 4pm closing time to
pick up their children (despite the contracted pickup time of
4pm).
• Although day-care staff frowned on this behavior, there was no
specific sanction for tardiness.
• The experimental manipulation was, at 6 of 10 day-care centers,
to fine parents for being late.
Effect of fine on late pick-up of pupils from
day-care centers (Gneezy and Rustichini)
• The procedure was:
– In the first 4 weeks, collect baseline data. no
manipulation
– In weeks 5 through 16, a fine was announced and
parents were charged approximately $3:50 for a
delay of 10 minutes or more.
– In weeks 17 through 20, the fine was removed
without explanation.
Effect of fine on late pick-up of pupils from
day-care centers (Gneezy and Rustichini)
• Results:
– In the first 5 weeks in which the fine was in place,
late arrivals approximately doubled at the
treatment centers. that is, the fine induced more
late arrivals.
– In week seven forward, late arrivals equilibrate at
slightly less than twice the initial level.
– After week 17, when the fine was removed, late
arrivals did not return to normal.
Effect of fine on late pick-up of pupils from
day-care centers (Gneezy and Rustichini)
• Interpretation:
– Incomplete contracts.
• The formal day-care contract did not specify penalties for
coming late, but it was implicitly understood that this was an
imposition on the staff and should only occur in emergencies.
• Because the consequences for violating the contract were not
specified, parents did not know how much they stood to be
punished for tardiness (i.e., would there children be asked to
leave the day-care?).
• The fine revealed to parents these consequences were very
mild. Once the consequences were known, many parents
decided they were happy to bear them.
Effect of fine on late pick-up of pupils from
day-care centers (Gneezy and Rustichini)
• Interpretation:
– Social norm.
• Prior to the fine, parents implicitly understood that arriving
late was taking advantage of teachers’ generosity, and so
refrained from doing so in compliance with a social norm.
• The fine turned this informal social interaction into a
market transaction. The fine conveyed to the parents that
they could buy the right to be late.
• Why did behavior not return to normal when the fine was
removed? Perhaps the perception of the social norm was
altered by the experience of the fine. Parents continued to
think of lateness as an additional service rather than a
violation of terms.
Effect of fine on late pick-up of pupils from
day-care centers (Gneezy and Rustichini)
• Interpretation:
– The authors state: “The conclusion seems to be that the
absence of a price is not just the limit of very low prices.
Mentioning a payment is enough to change the perception
of the contract: from a service, which is due from them as
subjects in the experiment, to a market exchange. In a
similar manner, in the day-care study a fine is enough to
change the perception of the obligation to arrive on time.”
– explicit incentives exist against a backdrop of implicit rules
and incentives (typically called social norms). By
introducing explicit incentives one may inadvertently
‘crowd-out’ or undermine implicit incentives.
Crowding out of monetary incentive
(Frey and Oberholzer-Gee 1997)