The relationship between
carbon dioxide emissions and
economic growth
Oxbridge study on CO2-GDP relationships,
Phase 1 results
Michael Grubb,1 Benito Müller,2 and Lucy Butler1
1 Department
2 Oxford
of Applied Economics, Cambridge University
Institute for Energy Studies and University of Oxford
Overview & datasets
ƒ Aims: to understand past relationships between
national CO2 emissions and GDP to help inform current
debates about emission projections
ƒ Datasets considered:
– International Energy Agency
– Energy Information Administration (US DOE)
– CDIAC (US Oak Ridge)
– WRI CAIT
ƒ No major inconsistencies observed, EIA accessible for
general trend analysis, CDIAC and CAIT for data since
1950, WRI CAIT most complete for cross-comparisons
ƒ Population is an important factors; all comparisons
analysed on per-capita basis
General finding: viewed across all data, basic
industrialisation clearly implies higher emissions but
link between wealth and CO2 is very weak beyond this
Per Capita Emissions vs Per Capita GDP, 2000
Carbon Emissions Per Capita,
(tonnes)
9
8
7
6
5
4
3
2
1
0
0
10000
20000
30000
40000
GDP per capita (1995 US$)
50000
60000
A. Time series patterns in industrialised
countries
For major countries which industrialised earliest – UK
& US - per-capita emissions have remained close to
levels 50 years before whilst per-capita GDP trebled
350
300
GDP
250
200
150
CO2
emissions
100
50
Per Capita Emissions (UK)
Per Capita GDP (UK)
Per Capita Emissions (US)
Per Capita GDP (US)
Data Source: US Energy Information Administration
98
19
95
19
92
19
89
19
86
19
83
19
80
19
77
19
74
19
71
19
68
19
65
19
62
19
59
19
56
19
53
19
19
50
0
In last two decades, wider groups of developed
economies show divergence between GDP and
emissions with no clear linkage with GDP variations
180.00
160.00
GDP
140.00
120.00
CO2
emissions
100.00
80.00
60.00
20
02
20
00
19
98
19
96
19
94
19
92
19
90
19
88
19
86
19
84
19
82
19
80
40.00
EU (Emissions)
EU (GDP)
US (Emissions)
US (GDP)
CANZA (Emissions)
CANZA (GDP)
Japan (Emissions)
Japan (GDP)
CANZA = Canada, New Zealand, Australia
Data Source: US Energy Information Administration
In the ‘post transition’ period, transition economies
(EITs) have grown with little or no emissions growth
160.00
GDP (Except
Ukraine)
140.00
120.00
100.00
80.00
60.00
40.00
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
Ukraine (Emissions)
EU Accession (Emissions)
Russia (Emissions)
EIT (Emissions)
Ukraine (GDP)
EU Accession (GDP)
Russia (GDP)
EIT (GDP)
B. Scatter analysis of CO2 – GDP
relationships in industrialised countries
For scatter analysis, care needs to be
taken to avoid spurious results
ƒ Oxbridge dataset for scatter analysis of industrialised
countries:
– Source data from WRI CAIT includes all present OECD and
EIT countries
– Data problems for some EITs prior to 1992
– Countries with population < 5m excluded to avoid small
county project and border effects (eg. Luxembourg steel
plant); Latvia, Lithuania and Estonia aggregated to Baltic
data
– Data examined subject to per-capita emission thresholds
of (a) 1tC/yr and (b) 2tC/yr to explore sensitivity to
degrees of prior industrialisation
– Data smoothed over (a) 3 years and (b) 7 years to
examine variational effects
Scatter analysis of data shows wide dispersion that
declines for longer averages and higher emissions
- only South Korea showed strong GDP-CO2 link at high growth rates
Per Capita GDP Growth vs Per Capita Emissions Growth
Growth Rate of Emissions pc
Carbon Emissions > 1 tonne, 3 yr Moving Average
Carbon Emissions > 2 tonne, 3 yr Moving Average
20%
15%
15%
10%
10%
5%
5%
0%
0%
-5%
-5%
-10%
-10%
-15%
-15%
-20%
-25%
-20%
-15%
-10%
-5%
0%
5%
10%
-20%
15%
Carbon Emissions > 1 tonne, 7 yr Moving Average
Growth Rate of Emissions pc
-10%
-5%
0%
5%
10%
Carbon Emissions > 2 tonne, 7 yr Moving Average
10%
10%
5%
5%
SOUTH
KOREA
0%
SOUTH
KOREA
0%
-5%
-5%
-10%
-10%
-15%
-15%
-15%
-15%
-10%
-5%
0%
5%
10%
-15%
-10%
-5%
0%
5%
10
Specific growth rate findings from scatter
analysis of OECD & EIT countries
ƒ
ƒ
ƒ
ƒ
There is wide dispersion of results, with per-capita emissions
growth in the dataset mostly in the range +/- 10% (3-year
trend) and +/-5% (7-year trend) about an average of almost
stable per capita emissions
There are short bursts of emissions growth exceeding 10%/yr,
but
the only cases with emissions growth sustained above 5%/yr
for 7 years or more are:*
– The ‘tiger economies’ in the earlier stages of their basic
industrialisation whilst emissions still less than 2tC/yr
– South Korea in the ten years after the oil price crash
Excepting South Korea, sustained (7-year avg) GDP growth
rates above 4% are associated with emission trends within
range +/- 3%/yr.
* The Danish moving average to 1996 (only) exceeded 5%/yr due to combination of oil
price collapse with extended low rainfall reducing imports of Scandinavian hydro power.
In the transition economies, resumed economic
growth in many of the fastest-growing economies has
been accompanied by continued emission reductions
GDP v s CO2 Annual Growth Rate s
(Pe r Capita, 1995-2001)
0.03
CO2 per cap Growth Rates
0.02
Belarus
0.01
Bulgaria
0
-0.01
Baltic States
Russia
Hungary
Poland
Moldova
Slovakia
Czech Republic
-0.02
-0.03
Ukraine
-0.04
-0.05
-0.02
Romania
-0.01
0
0.01
0.02
0.03
0.04
0.05
0.06
GDP per cap Grow th Rates
Note: there are uncertainties about the Belarus data
Data Source: US Energy Information Administration
C. A closer look at carbon intensity trends in
countries with economies in transition
Emission intensities trends in EITs were very varied during
recession periods (first half 1990s) but since 1996 most
have averaged 3-10% annual improvement
Annual rate of decline in Carbon Intensity (%/yr)
0.15
0.1
0.05
0
-0.05
1991-1995
1996-2001
Uzbekistan
Ukraine
Turkmenistan
Tajikistan
Slovenia
Slovakia
Russian Fed.
Romania
Poland
Moldova
Macedonia
Lithuania
Latvia
Kyrgyzstan
Kazakhstan
Hungary
Georgia
Estonia
Czech Republic
Croatia
Bulgaria
Bosnia
-0.15
Belarus
-0.1
The median rate of intensity reduction across all the
EITs was by late 1990s more than 4%/yr
95
96
97
98
Annual % rate of change of
carbon intensity (Carbon / GDP)
0.0%
-1.0%
-2.0%
-3.0%
-4.0%
-5.0%
EIT Median of Annual Percentage Changes
(Based on 5-year running average data)
99
Principal conclusions
Conclusions regarding GDP-CO2
relationships: General and OECD
ƒ
ƒ
ƒ
Data are highly variable over time and between countries; any
generalisations need to be treated with great care
Beyond basic industrialisation, any relationship between GDP
and CO2 appears to be very weak
The available historical data for OECD indicate:
– Major ‘early industrialisers’ (US, UK) show evidence of
saturation in per-capita emissions, but at very different
levels
– Excluding small country effects, no country other than
South Korea after oil price collapse has sustained per capita
emission growth rates (7 year average) above 5%/yr
– With this exception, since 1980 any link between emissions
and GDP appears very weak and this has not changed in
the period of energy price stability (1990+): there is no
clear link between more rapid economic growth and more
rapid CO2 growth
Conclusions regarding GDP-CO2
relationships: EITs
ƒ For Economies in Transition, comprehensive analysis is
constrained by inadequate data prior to 1992, but:
– The collapse of emissions associated with the initial
transition has not reversed during the period of subsequent
economic growth
– Countries that have implemented reforms have since mid
1990s experienced resurgent economic growth without
emissions growth
– For these countries, annual carbon intensity improvements
since mid 1990s have been in range 3 – 10%/yr with the
median across all EITs exceeding 4%/yr
Postscript: our findings in relation to the
Institute of Economic Analysis paper on
‘Economic Consequences of Possible Ratification’
ƒ
ƒ
ƒ
ƒ
The Institute paper presents extensive and interesting analysis of international data,
but appears quite selective in its use and questionable in the analogies drawn
Our findings on the tendency of many advanced economies towards approximate percapita emissions stabilisation is in sharp contrast
Our findings regarding intensity trends appear very different. Major reasons include:
– Institute analysis of intensity vs. economic growth (Fig. 31 and associated
paragraphs) is driven largely by the negative data (economic collapse of EITs) and
the results of early industrialising and OPEC oil exporters;
– OECD and EIT data (as in our analysis) address the experience of diversified and
growing economies with an established industrial base
– The Institute’s use of 2%/yr carbon intensity improvement is hard to reconcile with
evidence from other transition economies
The biggest issues are:
– all the evidence suggests that emissions are to an important degree a function of
policy and choice that determines the energy efficiency of economies
– no industrialised country other than South Korea after the oil price collapse (which
started from a far lower emissions and intensity base) has sustained the rate of
emissions growth projected in the Institute’s analysis over a comparable period
– Is the Russian economic structure analogous to OPEC or to other EITs, and do its
aspirations lie to a diversified OECD-like market economy, or a primary resource
economy with subsidised energy prices?