DG SANCO

1

Conference How Can Behavioural Economics
Improve Policies Affecting Consumers? Brussels,
28th November 2008 Thoughts on behavioural
economics and consumer regulations Matthew
Rabin, University of California, Berkeley
2
Conference questions

• Are we yet in a position to integrate this stuff?
• Does Behavioural Economics (BE) call for more
  interventionism? Less? Where?
• What next? Research and policy tools?

3
Questions for me

• Main insights from BE
• Can BE serve to complement survey data?
• Robust enough for application now?
• More for some markets than others?

4
The postulates of Neoclassical Economics

• Perfect competition
• Perfect information
• Perfect rationality
• (e.g., self-regarding preferences, time
  consistent preferences)

5
How neoclassical economics tenets were relaxed
over time

• 1930s perfect competition
• gt monopolistic and oligopolistic models
• 1970s perfect information
• gt asymmetric information
• 1990s perfect rationality
• gt bounded rationality

6
Nobel Laureates

• 2002 Kahneman
• "for having integrated insights from
  psychological research into economic science,
  especially concerning human judgment and
  decision-making under uncertainty"
• 2002 and Smith
• "for having established laboratory experiments
  as a tool in empirical economic analysis,
  especially in the study of alternative market
  mechanisms"

7
Behavioural Economics (BE)

• BE studies how people actually make choices
• It "increases the explanatory power of economics
  by providing it with more realistic foundations"
  (Camerer, 2003).

8
BE, main insights

• Many, but many specific errors, too
• More relevant discussing the underlying
  principles and the big emerging theme
• Ill illustrate with specific error bias in
  predicting future tastes

9
BE, main insights (2)

• Main principle for regulation help people
  achieve their own goals
• Take a set of options x, y, (e.g., say
  Adjustable Rate Mortgage and Fixed Rate Mortgage)
• Consider regulation when reason to believe
  consumers commonly choose x even though they want
  y.
• A first test any way to them to choose y? If
  people seem to want x no matter what, shouldnt
  be in business for promoting y.

10
BE, main insights (3)

• A useful framing of BE findings revealed
  preference fails because choice varies with the
  situation (see the Experiment carried out with EU
  Consumer Policy officials)
• Most (not all) behavioural phenomena in
  environment A people choose x from x, y in
  environment B, y from x, y, whereas
  conventional rational choice says choices should
  be the same. Present bias, framing effects,
  hot/cold gap, etc.
• Want borrelnoten now? Borrelnotten 7 days from
  now?
• Value for car while with salesperson? Next
  morning?

11
Preferences

• Must use something beyond traditional economic
  notion of preferences to find which choices, if
  any, reveal true goals.
• Let consumers choose simply not sufficient for
  evaluation of well-being
• Use BE principles to make guesses based on
  peoples behaviour and other information (as to
  what choice, across a variety of environments,
  best represent true preferences.

12
BE intervention

• General principle of BE intervention use light
  paternalism to affect when and how people choose
  from x, y, not limiting choice to y.
• Prospective choice better than immediate choice
  when there is present bias if I want candy or
  cocaine, make me choose ahead of time
• considered choice vs. empathetic choice

13
BE intervention (2)

• Letting people choose from x, y generally
  better than imposing y. This can accommodate
  heterogeneity in tastes and in degree of
  rationality.
• When done right, it can encourage those who
  should be doing y but mistakenly dont do so,
  while letting those who wisely choose x still do
  so.
• E.g., prospective choice and self-control making
  people who want to drink gin or eat candy at 2am
  decide before 11pm. Lets those who really want to
  plan for it, but those who dont really want to,
  not do it at the last minute.

14
Why intervention?

• Because market may deliver one way of choosing x
  vs. y, rather than what BE theories suggest best
  way of choosing x vs. y
• E.g., free market sells people things whenever
  want
• Tempting good consumed according to immediate
  preferences, not prospective preferences.
• Durable goods bought when most enthusiastic, not
  under true long-range average enthusiasm

15
Why intervention? (2)

• NOTE burden is greater than observing that
  market provides context A, but that a BE theory
  tells us B would be better.
• Must think through logic if our theories are
  really right, why wouldnt the market provide B?
• Specific class of error and intervention
  illustrates

16
Mispredicting Preferences

• 2 ways tastes change over time
• Temporary Fluctuations (moods, cues )
• Adaptation (paraplegia, standard of living),
  addiction.
• Evidence people under-appreciate (even very
  predictable) changes in their tastes, and hence
  falsely project their current tastes onto the
  future (e.g., shopping on an empty stomach leads
  people to buy too much).
• Lots of examples and relevant for consumer
  policy. Peoples preferences fluctuate.
• Fluctuations can be exogenous (e.g., random
  moods) or might be endogenous/market-induced
  (e.g., sales jobs, advertising )

17
Example of fluctuating preferences

• Nice simple experiment, dramatic context
• BBGJLM (2007) studied 13 long-time adult heroin
  addicts who had regularly receiving BUP, a
  medication
• Over an 8-week period, each subject was asked
  whether s/he would prefer each of 12 different
  amounts of money (ranging from 0 to 100) to a
  second dose, after receiving the initial dose.
• IC Subjects were told that one of their choices,
  randomly selected, would be implemented. Hence,
  they had the incentive to choose according to
  their true incentives.

18
Example of fluctuating preferences (2)

• Deprived (2 hours before scheduled dose) and
  satiated (rights after dose) addicts asked for
  their willingness to pay for the 2nd dose today
  or 2nd dose for 5 days hence.
• Average value of an additional dose
• When they would get the dose
• Today 5 days from now
• Current Deprived 75 60
• Craving Satiated 50 35
• These expressed willingness to pay are for the
  exact same circumstances by highly experienced
  addict. The rational choice would be identical
  prices.

19
Example of fluctuating preferences (3)

• 21 ratio of prices Willingness-To-Pay for BUP
  depending on when ask. Cant use the revealed
  preference approach.
• Which is the true preference?
• BE principles say it is probably the 35 that is
  most accurate reflection of her true experienced
  well-being.
• Not because drugs are bad and this is the lowest
  (present bias says 2nd column better than 1st,
  and projection bias says bottom row better than
  top
• What would unregulated market provide?
• Probably a combination of everything except the
  35!

20
More topical examples

• Consumers currently excited by a product, either
  because theyll enjoy first day, or successful
  salespersonship.
• Example of a class of regulation Cooling-off
  periods encourage persistent choice (for
  irreversible or hard-to-reverse decisions with
  long-term consequences, only allow implementation
  if made at more than one moment in time.
• Oversimplifying
• Markets consumers pay up to the highest WTP
• Right thing in many cases the lowest WTP
• E.g. Cooling-off / persistent choice
• Reminder for rational guys, lowesthighest, so
  little harm!

21
Final remarks

• PERSISTENT CHOICE is superior to either bans or
  enforcing all markets contracts. Marriage,
  suicide, durable goods, jobs, tattoos, etc.
• OF COURSE this may come at a cost. But when no
  serious cost to consumer or firm from delaying
  choice, little harm.
• CRUCIAL NOTE foregone sales, because consumers
  want to back out, does not properly count as
  economic cost to firm in this case this only
  happens when it is a sale that should not take
  place.

22
ThanksMatthew Rabin University of California,
Berkeley