Transcript commitment devices

Self-control
and the
Future of
Capitalism
‘My father rode a camel. I
drive a car. My son flies a jet
plane. His son will ride a
camel’
(Saudi epigram)
© Avner Offer
University of Oxford, 2010
1
The miracle of markets
• Consumer choice. Restrictive assumptions:
– Agent knows her own preferences; they are ranked
consistently
– All options known
– Choice voluntary, not influenced or coerced
• Individual choice aggregated by ‘invisible hand’ to
maximise welfare.
• A great deal is at stake.
– Economically efficient
– Individual sovereignty justifies liberal society.
– ‘Freedom’ not only instrumental, also a good in itself.
Validates dignity.
– Justifies policy priority of maximising ‘choice’
2
Choice over time
• Are conditions for rational choice satisfied? Not clear:
– Coffee or tea? Immediate choice is trivial
– Get married or walk away? Choice over time is
intractable. Not sure even in retrospect.
– Significant choice is over time, between now and future.
– How well can we do it?
Myopia gives rise to bad choices:
‘There is no quality in human nature, which causes more fatal
errors in our conduct, than that which leads us to prefer
whatever is present to the distant and remote.’
[David Hume, Treatise on Human Nature]
3
Myopia: an optical analogy
£120
£100
Derived from George Ainslie, Picoeconomics (1992)
time
4
Preference reversal:
Myopia, impatience, timeinconsistency
Validated in several disciplines
over five decades: animal
behaviour, psychology, economics,
neuroscience
£120
£100
time
5
Discounting: Present value of delayed payoff
200
180
Present
Value
160
140
120
V= present 100
value
A = Payoff 80
r = discount 60
rate
40
d = delay
•Exponential
A
V

discounting:
(r  1) d
Consistent &
normative
A
V

•Hyperbolic
d 1
discounting:
time-inconsistent &
positive?
200
100
20
0
15 14 13 12 11 10
9
8
7
6
• Calculation is genuinely intractable
Ainslie, Picoeconomics (1992); Good source: Loewenstein et al, Time and Decision (2003)
5
4
3
2
1
0
Delay 6
Discounting: The future is genuinely inscrutable
• It may not arrive
• No unique single time-preference.
– Different goods, different discounts, so no consistency
• Pizzas vs. pensions. Ice-cream vs. art.
• Experience is biased, so difficult to learn. (Khaneman, 2000)
• Unreliable prediction of wants (Gilbert, 2006)
• Neuroeconomics: Myopia ‘hard-wired’?
– Emotional incentives: desire, shame, anger, response to
provocation or stimulation. (Camerer et al, 2005)
– Short-term reward excites emotional brain locations, longterm excites ‘planning regions’ (McClure et al, 2004)
– Evolution: Myopia: core amygdala ‘reptilian brain’;
Foresight: calculating frontal lobes (Drewnowski, 2010)
7
An intertemporal dilemma
•
•
•
•
•
No maximising algorithm available
Myopia is our natural mode of discounting?
Choices still have to be made – but how?
Fall back on social conventions, norms, institutions.
Pre-existing, socially evolved, proven ‘commitment
devices’
• Commitment problem: How to sacrifice something
now, in return for something better later.
8
Overcoming myopia: commitment devices
• Personal strategies. Self-control:
– Attention control, personal rules, ‘bright lines’, hostaging.
• Social commitment technologies. Third party enforcement:
Counting, calendar, clock, Sabbath, money, gold standard, central
bank, law, constitution, contracts, education, examinations,
marriage, insurance, mortgages, pensions, commercial brands,
exponential discounting. ‘Civilization is a web of
commitments’
• Take time to evolve, develop, validate, diffuse.
– Clock:
• Public clock
• ‘Grandfather’ domestic clock
• Personal watch
• Alarm clock
• Snooze control
• Commitment is slow and costly to build up
• Role of government: commitment agent for individuals/society 9
Flow of innovation undermines commitment devices
• e.g. ‘family meal’ vs. foraging 24/7
Big Mac relative price
20
Mean body weight, UK & USA, BMI
SPA
32
BMI=kg/m2
SWE
GER
ITA
UK
UK
30
SWE
ITA
FRA
SPA
CAN
US
US
UK
US
ITA
CAN
US
US
AUS
USA Women
26
FIN
FIN
UK
SPA
NOR
GER
SWE
CAN
US
UK
FIN
NOR
GER
SWE
ITA
US
AUS
SWE
FIN
SPA
UK
FRA
GER
ITA
CAN
US
AUS
10
UK
CAN
AUS
US
US Black Wom.
USA Men
SPA
UK
GER
ITA
US
28
UK Men
UK Women
24
UK Army 25-9
UK
FIN
ITA
CAN
US
22
5
UK
SPA
SWE
FRA
UK
GER
ITA
15
FIN
20
1994
1996
1998
2000
2002
year
Fitted values
2004
1930 1940 1950 1960 1970 1980 1990 2000 2010
Big Mac relative price
10
Offer (2010)
Offer (2006), updated
Globalisation
• Assumptions:
– International trade and comparative advantage
– Self-correcting market equilibrium
– Factor prices change incrementally
– Incremental market adjustment
– Everybody better off
• Not always so smooth:
– World War I, Communism, Great Depression, World War II
• The future: incremental or shocked?
– Financial disarray
– Climate change (‘Global Warming’)
– Energy depletion (‘Peak Oil’)
– Food shortages
– Immigration pressure
– Shifting comparative advantage (rising BRICS)
– Ageing/fiscal crisis
– Natural resource scarcity (water, minerals)
– Political violence (terrorism, war, revolution, nuclear)
– Runaway technologies
11
Climate
Change
Shock:
Projected
Impacts
12
Stern Report, UK
2006
Liquid Fuel Shock
Thirteen forecasts of alloil production to 2030
US Energy Information
Administration (2009)
UK Energy Research Centre (2009)
13
Sensitivity of Oil Peak
to size of Reserves
UK Energy Research Centre (2009)
Climate change:
Emissions paths to
stabilise at 550ppm
CO2e
14
Stern Report (2006)
What to do about it?
‘Everyone talks about the weather,’ said Mark Twain, ‘but
nobody does anything about it.’
President Nasheed of the Maldives: ‘What we really need is a
huge social 60s-style catalytic, dynamic street action… ‘If the
people in the US wish to change, it can happen. In the 60s and
70s, they've done that…’
‘…My sense of the US is that a fair amount of them simply don't
believe in it’ (The Guardian, 29 May 2010)
So nothing will be done….
Or maybe tomorrow…
15
Reactions: (a) denial
• It’s not happening/I don’t want to know/something will turn up
• Disagreement is not metaphysical:
– E.g. ‘free markets’; Arsenal vs. Manchester U.
• Outcome objectively true or untrue. Time will tell. Soon.
• Risk is compelling, but falls short of certainty.
• Forms of denial:
– Rational denial: special interest, high time discount
– Emotional denial: myopic. Protecting comfort zone from the
unthinkable.
– Rational deniers mobilize emotional deniers against action
– Denials: Armenian genocide, WW2 Holocaust,
Vietnam/Iraq/Afghanistan, smoking & cancer, obesity, AIDS,
USA guns, Housing Bubble, Tea Party, Climate Change, Peak
Oil, Gulf oil spill, etc.
• Optimism is not prudent:
• Asymmetry of outcomes: Prepare for the worst, nothing happens—
16
small loss. Do nothing, worst happens—a calamity.
Reactions: (b) rationality
•
•
•
•
•
•
•
•
•
•
Rationality does not provide the solution.
Rationally framed as a cost-benefit problem.
If benefits exceed costs, action follows
‘What sacrifices should we make now to benefit our richer selves
later?’
Rational procedure does not prescribe a discount rate
Stern Report: sacrifice 1% of GDP now for gain of 5-20% of GDP
later. c. 1.4% discount rate.
Nordhaus: Skeptical. Higher discount rates, modest cuts now,
larger ones later. c. %7 discount rate. But market discount rates
imply short time horizons.
In rational time-consistent model, if we act later, we should be
acting now as well. Absence of current action is a diagnostic.
Failure to be acting now suggests that our actual model is not
‘rational’.
If not ‘rational’, problem is not whether it is worth acting, it is
how to motivate action. Requires a commitment device 17
Reactions: (c) commitment
• A prospective emergency. Which commitment device?
• National Security as a commitment device.
• Does not maximise net present consumer benefit
– Precedent: Cold War
• Nebulous threat.
• Indefinite timescale
• Untramelled resources
• Social consent
• Bipartisanship
• International alliances. Less calculating than trade
• What for? Objectives:
– Conservation/Invest in technofixes/lock in carbon taxes
– Prepare Plan B
– Military are already planning [USA, Germany]
• Problem: ‘National Security’ requires an enemy for motivation
• But the enemy is ourselves
• Shocks are external. Challenge is to master our intrinsic
18
myopia. Is it a problem we can solve?
Motivating action
• ‘Efficient’ procedure (exponential discounting) is cost-benefit
evaluation. Appropriate for social planners.
• Myopic choices (time-inconsistent) likely to be followed by
voters, tabloids, politicians, even planners in their private life.
• Mixed economy of discounting. How to integrate efficient
(normative) and myopic (positive) discounting?
• Sample problem: stop cheap holiday flights.
• Appears insuperable. Voters won’t stand for it.
• But time-inconsistency suggests a strategy as well.
• Available to responsible politician/social planner.
19
Hyperbolic discounting as commitment device
• Credibly pre-commit to restrict flights at point t=7?
• Easier than acting now.
• Hyperbolic discounting: Short-term sacrifice smaller, perceived longterm benefit greater.
100
62.3
present value
32.5
exponential 7%
hyperbolic k=0.3
10
t=7
1
0
10
20
30
40
50
time delay
20
Acknowledgments to Barnaby Marsh
• Exponential discounting for evaluation.
• Problem with time consistency: If we are not acting now, we won’t be acting
later either.
• Declining discount rates useful for policy. Can’t act now, but can act later.
• Is it worth sacrificing £1 now to get £10 in 50 years?
d-rate 1.5%
present value (discount factor)
100
d-rate 7%
80
certainty equivalent
(declining rate)
STERN
60
Value now of £10 in 50 years
£5.0
[WEITZMAN]
40
£2.6
20
NORDHAUS
£0.3
0
0
10
20
30
time delay
40
50
21