CO2 Emissions – Who Are Europe’s “Dirty Men”?

In his recent Emissions Reduction, Renewables and Recession post Euan Mearns made the following statement: “In terms of CO2 reduction (i.e. lack of it), Poland, Norway and Germany are the dirty men of Europe.” As we shall see Euan’s claim is broadly correct, but the success a country has had or not had in reducing its CO2 emissions is only one of a number of indicators that can be used to gauge its carbon dirtiness (or cleanliness). Here I combine five different ones to rank 26 European countries by carbon cleanliness/dirtiness in order to obtain a more broad-based perspective as to who the dirty men of Europe really are.

Some information on data sources and adjustments before proceeding. The CO2 emissions and energy consumption data I used are the 2014 values given in the 2015 BP Statistical Review, which include emissions from coal, oil and gas only, plus IMF population and World Bank GDP data. However, to make sure that the results were based on energy consumption in each country I made two adjustments:

  1. In countries that are net oil & gas exporters (Denmark, Netherlands, Norway, Russia) the CO2 emissions generated during production of exported oil & gas are included in national emissions totals. These emissions are not related to national consumption, so I removed them by subtracting 55kg of CO2 from each ton of oil equivalent exported (number from Statistics Norway). This makes less than two percent difference to emissions from Denmark (minus 0.1 million tons CO2), the Netherlands (minus 1.2 million) and Russia (minus 29.6 million) but reduces Norway’s 2014 emissions by 9.3 million tons, or about 20%.
  2. Having removed the emissions it becomes necessary to remove the exports. I did this by subtracting the value of its oil and gas exports from each country’s GDP ($4 billion from Denmark, $13 billion from the Netherlands, $100 billion from Norway and $350 billion from Russia, data from various sources). This reduced Norway’s GDP by 25%, Russia’s by 19% but Denmark’s and the Netherlands’ by less than 2%.

Now to the results:

Indicator 1: Per-capita CO2 emissions

This is probably the most commonly-used metric, with a country with low per-capita CO2 emissions scoring higher than a country with high per-capita CO2 emissions. Figure 1 shows per-capita emissions calculated by dividing BP’s 2014 CO2 emissions by the populations given by the IMF:

Figure 1: Country rankings, per-capita CO2 emissions

Romania wins and the Netherlands comes last. Among the larger countries France and Italy do well and Germany and Russia do not. The UK comes mid-pack.

Indicator 2: Tons CO2 emitted per ton of oil equivalent (TOE) consumed

This is a measure of a country’s “carbon intensity”, with countries that emit less CO2 in consuming a specified amount of energy scoring higher than countries that emit more. The results are shown in Figure 2:

Figure 2: Country rankings, tons CO2 emitted per ton of oil equivalent consumed

Now Norway moves into first place, followed by Sweden and with Poland bringing up the rear. France again does well and Russia moves into the middle. The UK slips a little but still edges Germany.

Indicator 3: Dollars of GDP per ton of CO2 emitted:

Reducing CO2 emissions requires money, so the amount of wealth generated per ton of CO2 emitted is another applicable indicator. Figure 3 shows the results obtained by dividing nominal GDP by CO2 emissions:

Figure 3: Country rankings, dollars of GDP (nominal) per ton of CO2 emitted

Switzerland wins this category easily, France continues to do well, UK moves into the top eight, Germany makes it to mid-pack and Russia sinks back. The bottom of the list is in fact made up entirely of former East Bloc countries, with Ukraine firmly in last place.

Indicator 4: Dollars of GDP per TOE of energy consumed

CO2 emissions can be skewed by conditions beyond a country’s control, particularly by the availability or otherwise of hydro resources. Dollars per ton of total energy consumed (again given in tons of oil equivalent) is a better measure of the overall efficiency of a country’s energy use. Figure 4 shows the rankings for this indicator:

Figure 4: Country rankings, dollars of GDP (nominal) per ton of oil equivalent consumed

Nothing much changes at the bottom of the rankings relative to Figure 3 but a good deal of shuffling goes on at the top. Switzerland easily retains first place but Norway drops from second to seventeenth – an example of how abundant hydro can reduce a country’s CO2 emissions even though its overall energy use may not be very efficient. Sweden falls from third to ninth and Finland and France both lose three places. Germany, however, gains five, and the UK, Italy, Ireland and the Netherlands three each.

Indicator 5: percent change in CO2 emissions 2005-2014

The amount by which a country has reduced its CO2 emissions is widely regarded as being the key indicator of how dedicated it is to the task of saving the world from the ravages of climate change, but it’s actually not very diagnostic. It discriminates against countries that already have high percentages of hydro, nuclear and/or firewood, percentage reductions can vary substantially depending on which stop-start years are chosen (I chose 2005 as the start year because that’s when the Kyoto Protocol came into force), economic downturns can have and have had a major impact and bizarre CO2 emissions caps (some East Bloc countries can double their emissions and still come in under their Kyoto targets) have also had a distorting influence. But here are the results anyway:

Figure 5: Country rankings, percent change in CO2 emissions 2005-2014

Coming out of nowhere to claim the gold in this category is Hungary, followed by Italy, and with the other four PIIGS countries also in the top ten. The UK is back in mid-pack along with France, Finland and Ireland. Germany and Russia are once more close to the bottom.

And Euan was right. Using this indicator and ignoring Russia and Belarus, Germany, Poland and Norway are indeed the dirty men of Europe.

Final rankings:

Above we have ranked 26 European countries for carbon cleanliness/dirtiness using five admittedly imperfect indicators, but by combining them we can hopefully smooth out the worst of the distortions and come somewhere close to the truth. Figure 6 shows the final rankings, obtained by normalizing the individual country scores in each category so that the leading country scores one and the others proportionately less and by summing the normalized values. The reader is left to form his or her own judgments:

Figure 6: Final rankings, all 26 countries

Finally, Figure 7 shows the results for the 21 EU countries. Twenty-one is a convenient number because it can be divided into three groups of seven, and I have used these three somewhat arbitrary groupings, depicted in traffic-light colors, to define the clean countries, the dirty countries and those in between:

Figure 7: Final rankings, 21 EU countries

According to these results Sweden is the clean man of the European Union and Poland by far the dirtiest. The inclusion of Belgium, the Netherlands and in particular Germany in with the dirty men is notable. It seems that the Energiewende still has a long way to go.

This entry was posted in Energy, Political commentary and tagged , , , . Bookmark the permalink.

15 Responses to CO2 Emissions – Who Are Europe’s “Dirty Men”?

  1. Chaam Jamal says:

    perhaps total methane emissions (converted to carbon equivalent values) from coal, oil, and gas production and transmission facilities should also be included.

  2. Figure 5 is still significant for Germany mind considering that many of countries above it were starting from a lower base already. If the Energiewinde was a true success, it would be higher.

    Is it possible to post such a graph for the countries on the start date of the Energiewinde (I assume 2002 as a policy though it existed earlier?) This might give a better measure of the EGW.

  3. Peter Lang says:

    There are many different data sources available and so many different ways to compare and rank countries by dirtiness, evilness and CO2 emissions per this that and the other. I set out to compare the figures and rankings in Figures 1 to 7 against the data from IEA and UNFCCC. I am not totally confident in the BP statistics and prefer to use IEA and/or UNFCCC when available.

    I began by looking at the charts in the GlobalCarbonAtlas You can select what data type, units, countries, regions and graphic view you want to display. You can produce many of the charts in Figures above. However, you cannot include and exclude exports and special analyses that Roger Andrews has don in hos analyses.

    From this I ranked 23 European countries by t/CO2 per capita and by kg CO2 per GDP. I did this for both territorial emissions and consumption emisisons.

    The results are below (you can go to the link above to do this and more yourself and display in may different ways).

    Territorial emissions, tCO2 per person
    Rank Country tCO₂/person
    1 Romania 3.9
    2 Hungary 4.4
    3 Lithuania 4.6
    4 Sweden 4.7
    5 Switzerland 4.8
    6 Portugal 5.1
    7 France 5.3
    8 Spain 5.9
    9 Slovakia 6.1
    10 Italy 6.2
    11 Bulgaria 6.2
    12 Denmark 6.7
    13 Greece 7.3
    14 Austria 7.4
    15 United Kingdom 7.6
    16 Poland 8.1
    17 Ireland 8.1
    18 Finland 8.7
    19 Belgium 8.9
    20 Germany 8.9
    21 Czech Republic 9.8
    22 Netherlands 10.0
    23 Norway 11.5

    Consumption emissions, tCO2 per person
    Rank Country tCO₂/person
    1 Romania 4.5
    2 Portugal 6.3
    3 Bulgaria 6.5
    4 Greece 7.0
    5 Lithuania 7.1
    6 Hungary 7.6
    7 Spain 7.6
    8 Poland 7.9
    9 France 8.3
    10 Netherlands 8.5
    11 Slovakia 9.2
    12 Sweden 9.3
    13 Italy 9.4
    14 United Kingdom 10.2
    15 Czech Republic 10.6
    16 Denmark 10.6
    17 Austria 11.0
    18 Germany 11.1
    19 Norway 11.1
    20 Ireland 11.8
    21 Finland 15.3
    22 Switzerland 15.5
    23 Belgium 20.2

    Territorial emissions in kgCO₂ per GDP
    Rank Country kgCO₂/GDP
    1 Switzerland 0.09
    2 Sweden 0.11
    3 Denmark 0.15
    4 France 0.15
    5 Norway 0.17
    6 Ireland 0.18
    7 Austria 0.18
    8 United Kingdom 0.20
    9 Italy 0.22
    10 Finland 0.23
    11 Spain 0.24
    12 Germany 0.24
    13 Belgium 0.24
    14 Netherlands 0.24
    15 Portugal 0.29
    16 Greece 0.39
    17 Hungary 0.41
    18 Lithuania 0.46
    19 Slovakia 0.53
    20 Czech Republic 0.70
    21 Romania 0.72
    22 Poland 0.76
    23 Bulgaria 1.34

    Consumption emissions in kgCO₂ per GDP
    Rank Country kgCO₂/GDP
    1 Norway 0.17
    2 Netherlands 0.21
    3 Sweden 0.21
    4 Denmark 0.23
    5 France 0.24
    6 Ireland 0.26
    7 United Kingdom 0.27
    8 Austria 0.28
    9 Switzerland 0.28
    10 Germany 0.30
    11 Spain 0.31
    12 Italy 0.33
    13 Portugal 0.35
    14 Greece 0.37
    15 Finland 0.40
    16 Belgium 0.55
    17 Hungary 0.69
    18 Lithuania 0.71
    19 Poland 0.75
    20 Czech Republic 0.75
    21 Slovakia 0.79
    22 Romania 0.85
    23 Bulgaria 1.39

    This was as far as I got before I got distracted onto other matters.

    • Peter: Thank you. Any chance you could carry this analysis further once you are undistracted?

      One thing I should probably have noted in the post is that a country’s progress, or lack thereof, towards meeting its emissions targets is calculated based on the emissions it produces domestically. Energy imports or exports are not taken into consideration. Thus the applicable indicators in your case are the “territorial” numbers, which for per-capita CO2 usage are within 10% of mine except in the cases of Austria, Belgium, Netherlands and Norway. I haven’t checked into the reasons for the differences.

      • Peter Lang says:

        Roger Andrews,

        Thank you. It would take me a while, and I am likely to be distracted on and off for some time, as of some unwelcome news yesterday. But I’ll still make short comments.

        There are many differences in approach that have to be decided.

        1. territorial or consumption (I say consumption is the correct measure and the only fair measure, but it’s impossible to price internationally)?

        2. CO2 or CO2-e (i.e. just CO2 or all 23 greenhouse gasses)?

        3. What emissions sources – e.g. all GHG emissions sources, just fossil fuels, just energy, just electricity, all except LULUCF? etc

        4. tonnes CO2-e per country, per GDP, per capita, per MWh, per GJ, per other?

        5. per GDP is per US$, per EUR, and by exchange rate or PPP?

        6. Kaya Identity factors?

        IEA figures consider only emissions from energy consumption.

        To provide input to and support for the UN process, the IEA is making available for free download the “Highlights” version of CO2 Emissions from Fuel Combustion.

        This annual publication contains, for more than 140 countries and regions:
        •estimates of CO2 emissions from 1971 to 2012,
        •selected indicators such as CO2/GDP, CO2/capita and CO2/TPES,
        •a decomposition of CO2 emissions into driving factors,
        • CO2 emissions from international marine and aviation bunkers, and other relevant information.

        The Global Carbon Atlas figures are from CDIAC data (I have found large errors in their data in the past, so suggest this source be used with caution, and carefully checked against IEA and/or UNFCCC).

        I’d suggest IEA for energy consumption only and UNFCCC are probably regarded as the most authoritative sources for CO2 emissions.

        • Roger Andrews says:

          What criteria to use indeed.

          After giving this matter some thought I’ve come to the following conclusions:

          1. A country’s level of carbon “dirtiness” or “cleanliness” is academic. The country is the way it is because it’s the way it is. Arguing over how it got there and what it might have done differently is unproductive. If the goal is to reduce CO2 emissions the only thing that matters is by how much the country reduces its CO2 emissions.

          2. When calculating emissions reductions only emissions over which the country has direct control (i.e. what you are calling territorial emissions) and which can be measured with reasonable accuracy should be counted. These include CO2 emissions from burning fossil fuels & cement manufacture, methane emissions from agriculture expressed as CO2e and minor GHG emissions (NOX, HFCs) also expressed as CO2e. (I assume here that we know how to convert methane etc. into CO2e). They also include emissions from burning biomass, which despite claims to the contrary is not a carbon-neutral source.

          3. They don’t include “carbon leakage” emissions – i.e. the CO2 emitted in producing all the manufactured goods the country imports from China. These are China’s problem, and they can’t be calculated accurately anyway. They also don’t include land use and land use change (LULUC) emissions. We don’t know enough about the carbon cycle to calculate them either.

          And I’m no big fan of the Kaya Identity:

          F = P*G/P*E/G*F/E

          F is global CO2 emissions from human sources
          P is global population
          G is world GDP
          E is global energy consumption

          It tells us that CO2 emissions depend on population, GDP, energy consumption and – CO2 emissions. Duh. It’s mathematically meaningless too. Canceling out terms gives F = F.

          • Peter Lang says:


            I am surprised by this and some previous seemingly dismissive comments. It doesn’t seem to set a good example. I disagree with some of your statements and assertions , but sense there’d be no point discussing them with you.

            However, I will respond to your comment about the Kaya Identity because it stuck me as unnecessarily rude and arrogant.

            And I’m no big fan of the Kaya Identity:

            F = P*G/P*E/G*F/E

            It tells us that CO2 emissions depend on population, GDP, energy consumption and – CO2 emissions. Duh. It’s mathematically meaningless too. Canceling out terms gives F = F.

            It’s either misquoted or a misrepresentation. You didn’t provide a link where you got it from so I can’t see why you made the mistake. It also makes me wonder if you understand unit analysis – i.e. the units on both sides of the equation must equate. For other readers not familiar with the Kaya Identity Roger Pileke Jr explained it in this paper as follows:

            Together the four factors of the Kaya Identity explain the various influences that contribute to increasing atmospheric concentrations of carbon dioxide, as follows:

            (1) Carbon dioxide emissions = population x per capita GDP x energy intensity x carbon intensity

            (2) P = total population

            (3) GDP/P = per capita GDP

            (4) Energy intensity (EI) = TE/GDP = total energy (TE) consumption/GDP

            (5) Carbon intensity (CI) = C/TE = carbon emissions / total energy consumption

            According to IEA (p20) Kaya Identity is:

            Kaya identity
            C = P (G/P) (E/G) (C/E)


            C = CO2 emissions;
            P = population;
            G = GDP;
            E = primary energy consumption

            Perhaps you’d like to explain to IEA why Kaya Identity is “duh”

            I don’t understand why you wrote this comment in such a way.

          • Peter:

            Your comment merits just two responses:

            1. The equation for the Kaya Identity that I gave is identical to the one you give.

            2. My name is Roger, not Arthur.

          • Peter Lang says:

            Your comment merits just one response:

            My Apologies for the wrong name.

          • Peter Lang says:

            I should add, the reason I find the Kaya Identity useful is because it help to clarify why, if we want to cut global GHG emissions at a fast rate, we must focus on the last term, i.e. “Carbon intensity of energy consumption”.

            The Kaya identity is commonly used as rates of change of each term; e.g.
            – population growth rate
            – GDP per capita rate of change
            – Energy intensity (per GDP) rate of change
            – Carbon intensity (of energy) rate of change.

            Economists look at the historical rates for these factors and can draw conclusions about what rate of change can be achieved for each in the future. I gave the example of Roger Pielke Jr’s application of this in the two links I gave in the first comment.

            I argue that policy that carbon mitigation policy cannot control population growth rate or GDP growth rate. All countries want to maximise GDP growth rate and many want to raise the population growth rate (e.g. Australia – we need continually increasing numbers o skilled human resources to develop our country and provide an ever increasing standard of living).

            Energy intensity (per GDP) is declining slowly (actually faster before we started trying to control the climate). Government policies cannot make a huge difference to the the rate of improvement in energy intensity. Therefore, this is not the place to be putting most of our efforts if we want to reduce global GHG emissions.

            That leaves us with “Carbon intensity (of energy)” as the Kaya Identity factor we should focus most of our efforts on.

            Franc demonstrated what could be achieved in a relatively short period of time. In two to three decades they reduced their carbon intensity of electricity (t/CO2/MWh) to 0.069 t CO2/MWh. That’s around 10% of Germany’s and 8% of Australia’s CO2 emissions intensity of electricity. The world could achieve that too (over about half a century) if we (led by the USA) did this:
            How to make nuclear cheaper than fossil fuels

  4. Willem Post says:


    In 2007, German CO2 emissions on the national territory were 959 million metric ton, import emissions were 528 million metric ton, totaling 1487 million metric ton

    Germany consumed on its territory 888 million metric ton and exported 600 million metric ton, totaling 1487 million metric ton.

    • Willem: You’re quite right that export/import emissions balances – or if you like “carbon leakage” – should ideally be allowed for when estimating a country’s carbon intensity, but as I noted in my response to Peter Lang above this isn’t taken into account in calculating emissions reductions, I suspect largely because it’s too hard to estimate.

  5. oldfossil says:

    Thank you for unblocking me. In your post a couple of weeks ago you compared GDP growth with renewable penetration. Obviously it’s easier for a country to improve its GDP to CO2 yield when its economy is growing strongly because proportionately more energy is going into production and proportionately less into overheads. I tried to get Tim Worstall of Forbes fame to take a look at that earlier post and give us the benefit of his insight, but regrettably he either has a blind spot or he regards anything climatic as too hot a political potato.

    • Peter Lang says:

      Obviously it’s easier for a country to improve its GDP to CO2 yield when its economy is growing strongly because proportionately more energy is going into production and proportionately less into overheads.

      This is not the main game. The main game is that we need to maintain global GDP growth as high as can be achieved sustainably while at the same time substituting low GHG emissions fuels and technologies for fossil fuel energy. It is not possible to achieve that by incentivising renewable energy. We need to transition to nuclear power as quickly as possible. Ad to achieve that we need to remove the impediments we’ve loaded on nuclear power that are making it far more expensive than it should be.

      Are you familiar with the Kaya Identity? This explains what factors must be addressed to reduce global GHG emissions. The one to focus on is the CO2 emissions intensity of energy (t CO2/GJ).

      Roger Pielke Jr. “Reality Check”

      See how to apply the Kaya Identity in Section 2 “Methodology of Evaluation”:
      Roger Pielke Jr., 2010 “An evaluation of the targets and timetables of proposed
      Australian emissions reduction policies

  6. Peter Lang says:

    O/T for this thread, this new post on WattClarity (Australia) might be of interest to Euan, Roger and some other readers.

    [Watt Clarity is run by Paul McArdle who is CEO of ROAM Consulting International (I think)]
    What would “demand” look like, in a “10x” intermittent scenario

Comments are closed.