This document provides an overview and summary of Eric Beinhocker's talk on escaping the last Malthusian trap through a revolution in carbon productivity. Beinhocker argues that neoclassical economics is ill-equipped to address the challenges of climate change and facilitating an economic transition to a low-carbon economy. He outlines the world's need to redo the Industrial Revolution in a sustainable way, transition to low-carbon growth while minimizing impacts on welfare, and use unprecedented policy measures to drive these changes. However, traditional economics cannot explain economic growth and change, or adequately assess the trade-offs between welfare, growth and decarbonization. A new framework is needed to navigate the final Malthusian trap of climate change

1.
iied
Escaping
the
Last
Malthusian
Trap
A
talk
by
Eric
Beinhocker
INET
Oxford
London
6
February
2014
Copyright
©
2014
Eric
Beinhocker
All
rights
reserved

4. The
last
Malthusian
trap
Today’s
discussion
Why
neoclassical
economics
is
the
wrong
tool
for
climate
change
A
complexity
economics
view
of
growth
Escaping
the
trap:
creaAng
a
revoluAon
in
carbon
producAvity

5. The
last
Malthusian
trap
Why
neo-­‐classical
economics
is
the
wrong
tool
for
climate
change
A
complexity
economics
view
of
growth
Escaping
the
trap:
creaAng
a
revoluAon
in
carbon
producAvity

6. Some
2.5
million
years
of
economic
history
(in
brief)
World
GDP
per
capita
1990
internaHonal
dollars
World
populaAon
Thousands
7,000
7,000,000
6,000
6,000,000
5,000,000
5,000
4,000,000
4,000
3,000,000
3,000
2,000,000
2,000
1,000,000
1,000
0
-2,500,000 -2,000,000 -1,500,000 -1,000,000 -500,000
0
0
500,000
Year
Source:
US
Census
Bureau
Historical
EsHmates
of
World
PopulaHon;
Kremer
(1993)
-2,500,000 -2,000,000 -1,500,000 -1,000,000 -500,000
0
500,000
Year
Source:
DeLong
(2005);
data
2.5
million
to
1
million
B.C.
extrapolated

7. The
Malthusian
trap
(circa
1000
BC
to
1800
AD)
Malthus
in
a
nutshell
A
Wages
Subsistence
C
B
PopulaHon
12
Stagnant
incomes
Global
income
per
person
(1800
AD
=
1)
10
8
6
4
Malthusian
trap
2
0
1000
BC 500
BC
0
500
1000
1500 1800
AD
900
Rising
populaAon
Thousands
800
700
600
500
400
300
200
100
0
Source:
Clark
(2007)
500
1000
1500
1800
AD
Source:
US
Census
Bureau
Historical
EsHmates
of
World
PopulaHon;
Kremer
(1993)

8. UnHl
1800
Malthus
…then
a
third
of
the
world
escaped
ruled…
Global
income
per
person
(indexed
1800
AD
=
1)
12
10
8
The
Great
Divergence
6
4
Industrial
RevoluHon
2
Malthusian
trap
0
1000
BC
Source:
Clark
(2007)
500
BC
0
500
1000
1500
2000
AD

9. …
but
with
an
unsustainable
growth
model
…
Changes
in
greenhouse
gases
from
ice
core
and
modern
data
400
350
RadiaHve
forcing
(Wm2)
CO2
(ppm)
300
250
10,000
5000
Time
(before
2005)
Source:
IPCC
AR4
WG1
(2007)
0

10. …
and
another
third
of
the
world
are
poised
to
escape
Annual
household
disposable
income
Thousands
RMB,
real
2000
Number
of
households
(millions)
2005
2015
2025
CHINA
200
and
above
1.0
3.4
8.2
100-­‐199
1.6
5.7
19.0
40-­‐99
112.6
8.8
71.4
25-­‐39
Less
than
25
214.1
75.7
74.2
107.5
54.1
57.8
Thousands
RMB,
real
2000
INDIA
1000
and
above
1.2
500-­‐999
10.9
90-­‐199
Less
than
90
Source:
McKinsey
Global
InsHtute
9.5
5.5
2.4
200-­‐499
3.3
33.1
55.1
91.3
101.1
106.0
74.1
94.9
93.1
49.9

11. We
face
our
final
Malthusian
trap
Peak at 550 ppm, long-term
stabilization 550 ppm
Annual
emissions
implied
by
Copenhagen
Accord
pledges
(Gt
CO2e)
Peak at 510 ppm, long-term
stabilization 450 ppm
70
Peak at 480 ppm, long-term
stabilization 400 ppm
65
Low range of pledges
60
55
50
Probability of
temperature
increase
under 2˚C
30
2.0˚C
70-85%
35
3.0˚C
40-60%
40
Expected
temperature
increase
15-30%
45
1.8˚C
25
20
15
10
5
0
2005
High range of pledges
2010
2015
2020
2025
2030
2035
2040
2045
2050
Source: “Taking Stock – Emissions Levels Implied by the Copenhagen Accord,” Project Catalyst, February 2010.

12. What
we
need
to
do
and
quesHons
for
economics
The
world’s
to-­‐do
list
Re-­‐do
the
Industrial
RevoluHon,
creaHng
a
sustainable
economic
system
TransiHon
to
a
low-­‐
carbon
economy
with
minimal
impact
on
welfare
and
growth,
especially
for
the
developing
world
Drive
the
above
with
policy
–
conduct
global
social
engineering
on
an
unprecedented
scale
QuesAons
for
economics
How
did
the
first
Industrial
RevoluHon
occur?
How
might
we
create
a
new
one?
What
are
the
interacHons
between
welfare,
growth
and
de-­‐carbonisaHon?
How
do
we
assess
the
trade-­‐
offs?
What
are
the
leverage
points?
How
do
we
avoid
unintended
consequences?
Preserve
individual
freedom?
Unfortunately
tradiAonal
economics
ill-­‐equipped
to
answer
these
quesAons

13. The
last
Malthusian
trap
A
complexity
economics
view
of
growth
Why
neoclassical
economics
is
the
wrong
tool
for
climate
change
Escaping
the
trap:
creaAng
a
revoluAon
in
carbon
producAvity

14. Neoclassical
economics
cannot
explain
key
characterisHcs
of
the
economy

16. The
economy
is
viewed
as
an
equilibrium
system

17. The
economy
is
viewed
as
an
equilibrium
system
but
such
a
system
cannot
grow
explosively,
create
novelty,
nor
spontaneously
self-­‐organize

18. And
such
a
system
cannot
just
‘crash’
–
as
ours
has

19. The
accidental
history
of
equilibrium
in
economics

20. Neoclassical
failure
#1:
Theory
of
growth
Y
(t)
=
Output
Cannot
explain
the
Industrial
RevoluHon
F
(K
(t)
,
A
(t)
*
L
(t)
)
Capital
Knowledge
Labour
No
connecHon
with
the
physical
world
Source:
Bolow
(1956),
Romer
(1996),
Nelson
(1996),
Daly
(1999)

21. Neoclassical
failure
#2:
Human
behaviour
Theory
doesn’t
match
real
world
behaviour
ExponenHal
discounHng
Hyperbolic
discounHng
Example
Society
spends
$1
billion
today
to
save
10
lives
per
year
in
perpetuity
Social
cost
of
capital
equals
5%
ExponenAal
answer
Cost
=
$4.76
million
per
life
saved
Source:
Axtel
and
McRae
(2006a),
(2006b)
Hyperbolic
answer
Cost
=
$1
million
to
$4
million
per
life
saved

22. Neoclassical
failure
#3:
Cost-­‐benefit
analysis
“Discount
ate!’
‘Discount
rrate!”
Prof.
William
Nordhaus
Lord
(Nicholas)
Stern
• Climate
uncertainty
has
fat
tails
with
power
law
scaling
• Cost-­‐benefit
analysis
typically
assumes
away
the
tails
• Would
pay
a
lot
to
avoid
catastrophe,
e.g.
Weitzman’s
‘Dismal
Theorem’
Source:
Stein
(2006),
Nordhaus
(2007),
Weitzman
(2007),
Barker
(2008)

23. Neoclassical
failure
#4:
Time
symmetry
Cost-­‐benefit
analysis
and
discounAng
assume
path
independence
and
Ame
symmetry
Samuelson
:
M
R
S
(τ,
τ’)
independent
of
C
τ’’
But
climate
effects
are
highly
path
dependent
and
largely
irreversible
on
human
Ame
scales
Source:
Arrow
and
Fischer
(1974),
Frederich,
Lowenstein,
Donohue
(2002),
Dietz
(2007)

24. The
last
Malthusian
trap
A
complexity
economics
view
of
growth
Why
neo-­‐classical
economics
is
the
wrong
tool
for
climate
change
Escaping
the
trap:
creaAng
a
revoluAon
in
carbon
producAvity

29. A
different
explanaHon
–
the
economy
is
a
‘complex
adapHve
system’
Complex
AdapHve
System
Many
interacHng
agents
and
organizaHons
of
agents
Designs
and
strategies
evolve
over
Hme
Macro
parerns
emerge
from
micro
behavior

30. A
paradigm
shis
TRADITIONAL
ECONOMICS
COMPLEXITY
ECONOMICS
Dynamics
Economies
are
closed,
staHc,
linear
systems
in
equilibrium
Economies
are
open,
dynamic,
non-­‐linear
systems
far
from
equilibrium
Agents
Homogeneous
agents
• Only
use
raHonal
deducHon
• Make
no
mistakes,
have
no
biases
• No
need
to
learn
Heterogeneous
agents
• Mix
deducHve/inducHve
decisions
• Subject
to
errors
and
biases
• Learn
and
adapt
over
Hme
Networks
Assume
agents
only
interact
indirectly
through
market
mechanisms
Explicitly
accounts
for
agent-­‐to-­‐agent
interacHons
and
relaHonships
Emergence
Treats
micro
and
macroeconomics
as
separate
disciplines
Macro
parerns
emerge
from
micro
behaviors
and
interacHons
EvoluHon
No
endogenous
mechanism
for
creaHng
novelty
or
growth
in
order
and
complexity
EvoluHonary
process
creates
novelty
and
growing
order
and
complexity
over
Hme

31. Long
history
of
evoluHon
in
economics
(and
vice
versa)
Problems
•
Driven
by
a
biological
metaphor
for
the
economy
•
Not
built
on
a
general
computaHonal
view
of
evoluHon

32. EvoluHon
is
a
search
algorithm
for
‘fit
order’
VARIATION
SELECTION
AMPLIFICATION
Create
a
variety
of
experiments
Select
designs
that
are
‘fit’
Amplify
fit
designs,
de-­‐amplify
unfit
designs
REPEAT

33. EvoluHonary
search
through
‘deducHve-­‐Hnkering’

34. Technologies
evolve

35. Economic
evoluAon
occurs
in
three
‘design
spaces’
Physical
technologies
Business
plans
Social
technologies

36. What
would
economic
evoluHon
look
like?
Non-linear wealth
creation
Increasing variety and
complexity
Spontaneous selforganization

37. But
we
cannot
avoid
the
Second
Law
of
Thermodynamics
–
economic
order
does
not
come
for
free

38. Understanding
the
“mother
of
all
complex
systems”
???

39. The
last
Malthusian
trap
A
complexity
economics
view
of
growth
Escaping
the
trap:
creaAng
a
revoluAon
in
carbon
producAvity
Why
neoclassical
economics
is
the
wrong
tool
for
climate
change

40. Industrial
revoluAons
are
producAvity
revoluAons
Physical
technologies
Business
plans
Social
technologies
Rapid
evoluAon
(e.g.
“Cambrian
explosion”)
Rapid
rise
in
producAvity

41. How
do
we
evolve
higher
‘carbon
producHvity’?
Kaya
idenAty
F
=
Anthropogenic
(CO2
emissions)
Carbon
producHvity
~
=
Source:
Beinhocker,
et.
al.
(2008)
p
*
GDP
per
capita
PopulaHon
1
e
*
f
g
+
*
e
*
Energy
intensity
of
GDP
Non-­‐energy
emissions
and
other
GHGs
f
Carbon
intensity
of
energy
≈
$GDP
CO2e

42. To
grow
the
economy
and
reduce
emissions,
carbon
producHvity
must
rise
10x
to
$7,300
per
tonne
by
2050
World
GDP,
US$
tn
(real
2000)
150
125
100
75
50
25
0
146
+3.1%
per
year
41
2000 2010 2020 2030 2040 2050
Global
emissions,
tCO2e
60
55
-­‐2.4%
50
per
year
40
30
20
10
0
2,000
7,300
10x
740
0
2000 2010 2020 2030 2040 2050
20
2000 2010 2020 2030 2040 2050
Source:
Beinhocker,
et.
al.
(2008)
Carbon
producAvity,
US$
(real
2000)/tCO2e
Carbon
producHvity
=
8,000
GDP
6,000
Emissions
+5.6%
4,000
/
per
year

43. If
emissions
are
capped,
higher
economic
growth
requires
higher
carbon
producHvity
Carbon
producAvity
required
to
reach
20
Gt
CO2e
by
2050
US$
(real
2000)/tCO2e
16,000
Annual
real
growth,
%
14,000
-­‐2
-­‐1
0
1
2
3
4
5
12,000
10,000
Base
case
forecast
8,000
6,000
4,000
2,000
Carbon
producAvity
required
870
1,300
2,000
3,100
4,700
7,000
10,500
15,800
0
-­‐2
-­‐1
0
1
2
GDP
growth
required
to
hit
20Gt
at
BAU
carbon
producHvity
growth
Source:
Global
Insight;
IPCC;
McKinsey
analysis
3
4
5
Forecast
GDP
growth
rate
2008-­‐2050,
percent
Without
carbon
producHvity
growth
need
to
shrink
economy
by
>-­‐2%
per
annum

44. If
we
capped
emissions
and
lived
at
today’s
carbon
producHvity,
there
is
not
much
we
could
‘afford’
*
Emissions
from
land
use
change
not
included
**
Based
on
10Gt/year
sustainable
emissions
and
future
populaHon
of
10
billion
people
Source:
McKinsey
analysis

45. A
carbon
producHvity
revoluHon
is
required
three
Hmes
faster
than
the
industrial
revoluHon
Index
Year
0
=
1
10
Carbon
producHvity
growth
required
2008–50
8
US
labor
producHvity
growth
1830–1955
6
4
2
0
0
10
20
30
40
50
60
70
80
90
100
110
120
130
Years
Source:
Beinhocker,
et.
al.
(2008)

46. But
no-­‐one
today
is
close
to
required
carbon
producHvity
Carbon
producAvity
2007,
177
countries,
all
GHGs
excluding
LULUCF
Adjusted
for
purchasing
power
parity,
2050
target
=
$13,300
GDP/tonne
Carbon
producHvity
US$
000
(PPP)/
tCO2e
5.5
5.0
Saint
Kirs
and
Nevis
Switzerland
MauriHus
4.5
Bangladesh
4.0
Sweden
Sri
Lanka
Norway
Average
carbon
producHvity
3.5
France
3.0
2.5
Japan
Austria
2.0
Indonesia
Turkey
Germany
Mexico
Brazil
South
Korea
Iran
Australia
Venezuela
Saudi
Arabia
1.0
Nigeria
0.5
Liberia
0
Canada
United
States
Qatar
South
Africa
Russia
0
5
Turkmenistan
China
10
15
G8+5
Singapore
Italy
Pakistan
1.5
India
United
Kingdom
20
25
30
35
Prosperity
GDP
per
capita
US$
000
(PPP)
Source:
WRI
CAIT;
UNFCCC;
Global
Insight;
McKinsey
analysis
40
45
50

47. Carbon
producHvity
has
increased
over
Hme,
but
not
nearly
quickly
enough
*
5-­‐year
running
average.
Emissions
data
includes
CO2
from
fossil
fuels
and
cement,
with
projecHons
for
CO2
from
land
use
changes
and
five
non-­‐CO2
gases
(CH4,
N2O,
HFCs,
PFCs,
and
SF6)
Source:
IEA,
CDIAC,
OECD,
EPA,
CEC,
World
Bank,
US
Bureau
of
Economic
Analysis,
McKinsey
analysis

48. Technology
will
help
–
but
we
need
to
accelerate
innovaHon
and
buy
Hme
Source:
Farmer,
et.
al.
(2013)

49. Some
hypotheses
for
climate
policy
• Climate
change
is
far
riskier
then
convenHonal
models
lead
us
to
believe
– Fat
tails,
irreversibility,
path
dependence,
etc.
• Carbon
prices
may
be
necessary
but
not
sufficient
– EffecHveness
of
price
signals
in
noisy,
complex
markets
– Industrial
revoluHon
not
triggered
by
spike
in
labour
costs
alone
–
broad
socioeconomic
phenomenon
• Need
to
broadly
change
the
“fitness
funcHon”
of
the
economy
– RegulaHon,
standards
(e.g.
consumer
and
worker
safety
laws
early
20th
c.)
– Behaviour,
social
norms
(e.g.
slavery,
smoking)
• Policy
and
poliHcs
for
homo
realitus
vs.
homo
economicus
- The
revenge
of
poliHcal
economy
and
human
behaviour

50. Some
hypotheses
for
climate
policy
(cont.)
• Social
technology
innovaHon
just
as
important
as
physical
technology
– InsHtuHons
(e.g.
green
banks?)
– Laws
(e.g.
carbon
fiduciary
responsibility?)
– InformaHon
(e.g.
climate
risk
disclosure?
GDP
measures?)
• Must
accelerate
evoluHonary
innovaHon
process
– VariaHon
–
dramaHcally
increase
shots
on
goal
– SelecHon
–
bias
fitness
funcHon
toward
low
carbon
– AmplificaHon
–
capital
and
talent
flows
to
low
carbon
– CreaHng
green
innovaHon
clusters
• We
need
to
buy
Hme
for
tech
progress
- Role
of
natural
gas
as
bridge?
• InternaHonal
cooperaHon
needs
to
emerge
borom-­‐up
rather
than
top-­‐down
– EvoluHon
of
trade
regime
vs.
“Rio
Dream”
and
Copenhagen

51. Summary
Industrial
RevoluHon
enabled
a
third
of
the
populaHon
to
escape
the
Malthusian
trap
of
poverty,
hardship
and
disease
But
it
created
our
next,
and
possibly
last,
Malthusian
trap
–
climate
change
Escaping
that
trap
will
require
a
low-­‐carbon
revoluHon
on
the
scale
of
the
Industrial
RevoluHon,
but
at
three
Hmes
the
speed
Economic
revoluHons
are
profoundly
disequilibrium
phenomena
–
not
explained
well
by
neoclassical
theory
A
complex
systems
view
helps
us
understand
the
evoluHonary
processes
that
drive
disconHnuous
innovaHon
and
growth
Climate
policies
should
acHvate
and
leverage
economic
evoluHonary
processes
–
policymakers
need
new
ideas,
there
is
much
work
to
do!

52. ‘We
cannot
solve
problems
by
using
the
same
kind
of
thinking
we
used
when
we
created
them.’
ALBERT
EINSTEIN
Unless
we
truly
understand
the
system
we
are
dealing
with
we
will
fail
We
cannot
afford
to
fail
But
if
we
can
more
deeply
understand
that
system,
we
just
might
succeed