Experimental Economics, Environment and Energy - Lecture 5 ... · Consumer as a rational decision maker Budget constraint Consumer = budget constraint + preferences Budget constraint

Experimental Economics, Environment and EnergyLecture 5: Consumer behavior: rationality, biases & nudges

Paolo Crosetto

Lecture plan

1. Consumer as a rational decision makerI AxiomsI ConsequencesI Expected behavior

2. Consumer as a human beingI Biases: predictable deviations from rationalityI (several) examples

3. Choice architecture & behavior changeI nudgeI marketing

4. Applications to energy

Paolo Crosetto Experimental Economics, Environment and Energy

Consumer as a rational decision maker

Budget constraint

Consumer = budget constraint + preferences

Budget constraint

I Consumers have limited resources: they face a budget constraintI This means that they face trade-o�sI i.e. not all is available, typically there is more on o�er that the amount a

consumer can buy



(Rational) preferences

Consumer = budget constraint + preferences

Preferences

I Preferences are the economic equivalent of consumer taste. Economistsassumes they are �xed � you are sort of born with them.

I Formally, having preferences means that, for any pair of goods A, B:

Completeness The consumer can say if A � B (read: A is preferred to

B), or B � A or A ∼ B (read: is indi�erent to).Transititvity If A � B and B � C , then A � C

Local non-satiation For any bundle that the consumer buys, we can �ndanother bundle that the consumers likes even more.

Independence If A � B, then adding a third alternative C cannot makethe consumer invert her choice (and hence B � A)

These are the rationality axioms



Completeness

Completeness Just assumes that you know what you like and dislike �else you should know that you are indi�erent.

I Do you prefer pasta or pizza?

I A trip to Paris or to Madrid?

I A trip to Paris or a pizza?

I An iphone or a samsung smartphone?

I (but also): a day with your family ten days from now or three days withyour family now?

I A day with your family now or 10 shares of Apple, inc.?

I two hour of master's class or one hour exam?

I ...and so on, for any couple of goods



Transitivity

Transititvity Just asks you a minimal level of consistency: you cannotenter loops in which you are money-pumped

I Do you prefer pasta to pizza?I Do you prefer pizza to kebab?I then you must prefer pasta to kebab.I Else, you can be money-pumped!



Local non-satiaiton

Local non-satiation Just assumes that you canot have a global satiation

point. I.e.: you will always want something more

I Even if this seems absurd...

I Try to imagine you are the richest person on the planet. You have it all.

I Still, surely there is something you want more or in exchange or you wantback or...



Independence

Independence assumes that if I give you an irrelevant alternative, youshouldn't change your order of preferences

I Do you prefer pasta to pizza?

I If I add kebab in the choice set, you should still prefer pasta to pizza.



Consequences

Several consequences of these axioms exist

I Consumer choice should be stable over time

I It should be impacted only by relevant additional information (e.g. gettingto know that smoking causes cancer)

I Consumers shoudl maximise their well-being (utility) subject to the budgetconstraint

I hence resulting in optimal consumption at all times (or as much as currentinformation allows)

I If consumers can be described as rational,

I then all usual Micro 101, 102, 201... consequences apply: substitution andincome e�ects, downward-sloping demand functions, market equilibria,perfect competition, etc...

But consumers happen to be irrational


Consumer as a human being

(predictably) irrational

(real) consumers deviate from the axioms in predictable ways

I That is, not only they make mistakes (which would be OK: randommistakes on average cancel out)

I But they consistently make mistakes, in precise directions, and accordingto one or another precise rule

I The rules followed by consumers to simplify the choice problems are calledheuristics

I consistent deviations from the rationality axioms are called biases

In the rest of this lecture we will cover some important biases and theirpossible application to energy


Attraction e�ect

Regularity

One consequence of the Independence axiom is that...

This is called regularity and it is an aggregate property of markets


Attraction e�ect

But does it always hold?

Please choose


Attraction e�ect

But does it always hold?

Please choose again


Attraction e�ect

Subscribing to the Economist

Please choose


Attraction e�ect


Please choose again


Attraction e�ect


Dan Ariely, Predictably Irrational

I When presenting only two choices

Option Online only Online & Print Print onlyPrice 59 125 �Choice % 68% 32% �

I When presenting all three choices

Option Online only Online & Print Print onlyPrice 59 125 125Choice % 16% 84% 0%


Attraction e�ect

The decoy/attraction/asymmetric dominance e�ect


Attraction e�ect

ADE

This is called the Asymmetric Dominance E�ect.

Adding to a choice set an asymmetrically dominated option � that is, an option that is

dominated by some but not all the alternatives in the set � increases the choice share

of the now-dominant option, at the expense of the others.

Some terms:

Target the asymmetrically dominant option

Decoy the asymmetrically dominated option

Competitor the other option


Attraction e�ect

Why is this a problem?

ADE is a violation of the Independence to Irrelevant Alternatives axiom ofrational choice.

Under I.I.A, if in the set

{target, competitor}⇒competitor < target,

then in a set{target, competitor , decoy}⇒target � competitor .

At the aggregate level, this implies regularity

Pr(target){target, competitor} ≤ Pr(target){target, competitor , decoy}

That is, preferences are context-independent: changing the choice set shouldnot a�ect choice


Attraction e�ect

Evidence for ADE

I The ADE has been found in product choices among products:I beer 6-packs (quality vs. price) [Huber et al.]I cars (ride quality vs. gas mileage) [Huber et al.]I restaurants (distance vs. quality) [Huber et al.]I dates (good looking, bad looking twin, other) [Ariely et al.]I televisions (resolution vs. durability) [Pan and Lehman]I apartments (size vs. location) [Pan and Lehman]

I Herne also found ADE in political opinions in Finland

I Curiously, the e�ect has been observed in animals (honeybees, gray jays:Sha�r et al)


Attraction e�ect

ADE: why?

I dominance gives at least one reason to choose when uncertain: you knowthat the target is at least better than the decoy

I the choice might be complex, and you have an easy way to simplify it:look for dominance

I cognitive: similar (but dominant) things are seen bigger thannot-dominant ones (because they are easier to compare)


Default bias

The power of the default

% of consent to organ donation across countries

How to explain this? Culture? Economic development? Religion?this is due to the default option (opt-in vs. opt-out [Johnson and Goldstein,Science 2003])


Default bias

Default bias: why?

I inertia and laziness : the status quo is the choice that takes the leastamount of e�ort

I uncertainty: when we are not sure what to do and lack expertise in thearea in question, we consider the default as a form of advice

I loss aversion: switching away from the default requires a willful action: itmight bring to mind what we could lose by switching, which may make usreluctant to change

I procrastination: I'll do it for sure tomorrow...


Fixed vs. variable tari�s

Rational decision makers and timed consumption

Imagine a phone subscription

I Do you prefer to pay day-by-day according to your consumption (sms,calls, internet)

I Or to have a �at fee?

think again: do you really use up your �at fee? If not, then you are paying not

ot use your phone.



Paying not to go to the gym

I Study of data from three American GymsI Type of tickets:

I Single entry 12$I 10 entries 100$ (10$ each)I Monthly fee 85$I Yearly fee 850$

I Cancellation policy:I Single and 10x no cancellationI Monthly: need to cancel by the 10th of the month, else pay next month as

wellI yearly: automatically cancels at the end of the year

A rational decision maker should go for monthly only if he visits at least7 times a month; monthly gives the freedom to opt-out should one fail todo so, so we should see things adjust after a few months.



Paying more to go less!



Paying not to go to the gym: main results

I Users who choose (...) a �at-rate contract pay a price per averageattendance of over $17 in the monthly contract and over $15 in the annualcontract.

I The average forecasted number of monthly visits, 9.50 (s.e. 0.66), is morethan twice as large as average attendance, 4.17.

I On average, 2.31 full months elapse between the last attendance andcontract termination for monthly members, with associated membershippayments of $187.

I The survival probability after 14 months for the monthly contract is 17percent higher than for the annual contract.



Paying not to go to the gym: why?

I Risk aversion

I Overcon�dence over future attendance

I Procrastination / default bias for monthly ticket holders


Social norms

Social norms

Choices do not happen in a vacuum! Others are around

I You might care about what the others think of you [social conformity]

I You might care about what example you want to set

I You might want to be at least as good as someone else [keeping up withthe Jones's]

I You might want to keep your face, show high morals, be a good citizen[social norms vs. market norms]

I You might have little information, and use other people's choices as cues[information-driven conformism]

I You might just like to do like the others do [preference-driven conformism]

I ...


Social norms

Social norms vs. market rules

I School close at 4pm, but parents are frequently late

I Delays are costly for the school: need to pay teachers, etc

I what can be done?

Experiment (Gneezy et Loewenstein 2000)

Control group

I Observations over 20 weeks

I No intervention

TreatmentI 4 weeks: no intervention

I 12 weeks: �ne of 10 NIS (3 euro)

I 4 weeks: no intervention


Social norms

Results, I


Social norms

Results, II

The �ne has pushed the numbers of parents arriving late up. Why?

A �ne is a price

I Crowding out: l'argent déplace la norme socialeI On passe de 'il faut le faire'...I ...à 'je vais le faire mais c'est OK: je vais payer'

I Information: l'amende est un prixI Avant l'amende, les parents avaient des croyances relatives à l'ampleur du

cout généré par leur retardI L'amende (qui n'est pas énorme, à 3 euros) donne une valeur au retardI Cette valeur est assez basse: on peut donc bien croire que notre temps à

nous vaut plus que ça

Take-home: faites payer assez, ou ne faites pas payer du tout


Social norms: keeping up with the Joneses

Keeping up with the others

Many peple care about being not too di�erent from the others

I Especially upwards: if everyone has a sporty nice car, you want one too

I Keeping up witht he Joneses: keep your social status near to the one ofthe people you want to be associated to

This can (and has been) used to move people towards lower energyconsumption

I Opower in the UK: compare consumption to the one of othershttps://goo.gl/G4FyRg


https://goo.gl/G4FyRg

Behavioral change and nudges

Exploiting biases to nudge change

These (and other!) biases exist. Can they be exploited for policypurposes?

I The list is long: https://en.wikipedia.org/wiki/List_of_cognitive_biases

I they have largely been used by marketing and advertisement

I could they be used for policy?

Biases can allow a new policiy instrument

I Old: prohibition

I Old: tax and regulate

I Old: cap and trade

I New: nudges


https://en.wikipedia.org/wiki/List_of_cognitive_biases


Nudging

'soft paternalism'

I A nudge is not a formal regulation but a small change in the context, thesetup, the choice environment that makes people change behavior

I It is often not perceived as limiting the freedom of choice in a formal way� it just exploits our biases for policy purposes

I the 'choice architect' can 'nudge' choice towards a desirable outcome



Some famous examples

I Organ donation defaults

I Cafeteria position of meals

I Opower 'see what the others do' energy meter

I Study in the UL (London) about gas / energy consumption and socialnorm nudging (moodle)

I To incentivise public transport: lottery but not for car riders (regret)

I ...

Have a look at the list paper on moodle!



Some interesting nudge links

I French national competition for nudges: https://goo.gl/d7hNOf

I The Guardian nudge article: https://goo.gl/vYGTDN

I Telegraph nudge article: https://goo.gl/PeIyGa

Have a look at the list of nudges - pdf on moodle!


https://goo.gl/d7hNOf

https://goo.gl/vYGTDN

https://goo.gl/PeIyGa

Experimental Economics, Environment and Energy - Lecture 5 ... · Consumer as a rational decision maker Budget constraint Consumer = budget constraint + preferences Budget constraint

Documents