Assessing decoupling of greenhouse gas emissions from economic activity in the Netherlands: A 1990-2022 analysis

The Paris Agreement emphasizes the need to reduce GHG emissions to limit global warming to well below 2 degrees Celsius. Many countries have formally committed to these goals, and the Dutch government has integrated them into its national climate policies, aiming to reduce greenhouse gas emissions by at least 55% by 2030 compared to 1990 levels.[1] However, emissions are believed to be closely linked to economic activity and output (see appendix 1). The European Commission supports programs such as the European Green Deal to achieve these goals while stimulating economic activity. The key question is whether GHG emissions can be decoupled from economic activity in the long run.

Our analysis focuses on the Netherlands and provides insights into the factors that drove decoupling of emissions from the economy. We begin by evaluating aggregate Dutch GHG emissions and their relationship with economic activity, defined as real GDP, which is the sum of the value added of all goods and services produced in an economy, adjusted for inflation. Our findings indicate relative decoupling in the Netherlands, where GDP growth is accompanied by a slower increase in GHG emissions from 1990 to 2022. After adjusting for short-term economic fluctuations, we found evidence of absolute decoupling of trend GDP from GHG emissions, which indicates that during this period, GDP grew alongside a reduction in GHG emissions.

When examining different types of gases, we observed that there appears to be absolute decoupling between economic activity and emissions of methane (CH4) and nitrous oxide (N2O). While finding absolute decoupling for CO2 is more challenging, excluding biomass-related emissions makes its case stronger. Trends in emissions of all three gases appear to be primarily influenced by regulations and laws, suggesting the crucial role of government in shaping future directions.

Despite these positive trends, much more effort is needed to meet the Paris climate goals, a finding consistent with the most recent assessment by the Netherlands Environmental Assessment Agency (PBL) in their Climate and Energy Outlook (KEV) (in Dutch). Comparing current reductions to what is needed to meet the Paris goals shows that decoupling achieved so far is insufficient. Furthermore, our study focuses solely on production-based emissions. More research on consumption-based emissions is needed to provide further insight into the Dutch contribution to global climate change.

[1] The long-term goal is to achieve climate neutrality by 2050, which means net-zero greenhouse gas emissions (Dutch goals within the EU | Climate change | Government.nl)

Literature suggests that while economic growth and GHG emissions have historically gone hand in hand, this relationship may have changed over time (see appendix 1). However, studies on this relationship have produced varied results, depending on the choice of data, metric, country, time-horizons, and methods.

This report aims to answer two central questions: What do we find specifically for the Netherlands regarding decoupling of GHG emissions from GDP? Additionally, to what extent does decoupling occur across different types of GHG?

To answer these questions, we made several choices for our analysis based on the review of literature. We explain the details of these choices in appendix 1.

Due to the above setup, our analysis differs from PBL’s KEV in several ways. First, our analysis focuses on historical data, while KEV looks at future developments and policies that shape emission trajectories. Second, we explicitly consider economic variables in our analysis, unlike KEV, which focuses on emission processes. For this research, we use a measure of GHG emissions that aligns more closely with national accounts, therefore more suitable for environmental-economic analyzes. This is different from the measure used by KEV, which follows the Intergovernmental Panel on Climate Change (IPCC) guidelines for monitoring international agreements. Finally, we examine the overall performance of the Netherlands as a country and look into specific sectors only if it helps explain major changes in emissions of a particular GHG from 1990 to 2022. In doing so, we follow the NACE sector classification,which can be directly linked to the GHG emissions data we used, while KEV focuses on climate sectors – electricity, industry, built environment, mobility, agriculture and land use.

Figure 1 provides an overview of the development of GHG emissions in the Netherlands from 1990 to 2022. A few observations are worth making:

To compare with 1990, we converted the data to indices, setting the value for 1990 as 1.

In Figure 2 (left), we observe diverging trends in GDP and GHG emission levels in the Netherlands between 1990 and 2022 (both data series are indexed to 1990=1). In terms of growth rates, GDP growth outpaced emissions growth (relative decoupling) in most years during this period, and in one third of this period, GDP grew while emissions declined (absolute decoupling) (see figure A.1 in appendix 2). Regressing the growth in GHG emissions on the growth in GDP over the period from 1990 to 2022 results in an estimated elasticity of 0.27, meaning that on average, a 1% growth in GDP is associated with an 0.27% growth in GHG emissions during this period. Since the elasticity is between zero and one, it indicates relative but not absolute decoupling. In other words, while both GHG emissions and GDP grew, GDP growth outpaced GHG emissions growth during this period.

These results can be misleading, due to the influence of business cycle effects on both GDP and GHG emissions data. Our focus, however, is on the long-term relationship between GHG emissions and economic activity. Therefore, following Cohen et al. (2018), we use a Hodrick-Prescott filter[2] (HP-filter hereafter) to decompose emissions and GDP into their trend and cyclical components (see appendix 2 for methodology), and use the trend component to estimate the trend elasticity. Estimates using filtered data (see figure 2, right) reveal a statistically significant trend elasticity of -0.29 (see appendix 3), meaning that on average, a 1% increase in GDP is associated with a 0.29% reduction in GHG emissions. Hence by using the filtered trend component to focus on long-term trends, we observe absolute decoupling, where economic growth has been achieved alongside a reduction in GHG gas emissions in the Netherlands. This contrasts with the relative decoupling suggested by the unfiltered data, where emissions still grow but at a slower rate than GDP.

[2] The Hodrick-Prescott (HP) filter is a statistical procedure used to remove short-term fluctuations and highlight the underlying long-term trend.

Figure 1 shows that GHG emissions in 2022 were 44 megatons lower than in 1990. Figure 3 breaks down these developments by specific greenhouse gases: While carbon dioxide (CO2) accounts for the biggest share of total GHG emissions, only about 12 megatons of the 44-megaton reduction came from reduction in CO2. Additionally, for most of the period from1990 to 2022, CO2 levels remained above their 1990 levels, only falling below them for the first time in 2020. Over half of GHG emissions reduction between 1990 and 2022 came from lower emissions of methane (CH4) and nitrous oxide (N2O, also known as laughing gas), with reductions of 18 megatons and 9 megatons, respectively. Therefore, the overall decrease in Dutch GHG emissions during this period was primarily driven by reductions in non-CO2 emissions. This finding aligns with a report from the National Institute for Public Health and the Environment (RIVM) on Dutch emissions from 1990 to 2019.

To analyze the long-term relationship between different types of GHGs and GDP growth, we used the same approach (HP filter) and filtered trend data series (see figure 4). With all GHGs, there is a clear indication of relative decoupling from 1990 to 2022. However, regarding absolute decoupling, we only found statistically significant negative trend elasticities for CH4 and N2O (see appendix 3).

Why do different types of GHGs have varying relationships with GDP growth? In the following sections, we will examine GHG emissions by gas type and explore the factors that have influenced the relationship between emissions and economic activity.

CO2 emissions only modestly decouple from GDP: The role of biomass

CBS reports GHG emissions (in Dutch) of total CO2 emissions into two categories: CO2 excluding emissions from biomass and CO2 emissions from biomass. A comparison of the emissions from 1990 to later years shows that CO2 emissions excluding biomass peaked in 2004 and have been declining since 2010 (see figure 5). In contrast, biomass emissions have been increasing over time. Notably, since 2020, non-biomass-related CO2 emissions have fallen below their 1990 levels, while biomass CO2 emissions have risen sharply. This has resulted in a much smaller overall decline in total CO2 emissions.

A sharp rise in biomass emissions mainly comes from the energy and waste sectors

Biomass, often considered a renewable source of energy (bioenergy), has seen increased use in Europe, including the Netherlands. The most significant increase in biomass-related emissions is observed in the energy sector (see figure 6). This is likely due to the Dutch government’s subsidies through the SDE++ 2023 (Stimulation of Sustainable Energy Production and Climate Transition) program, which incentivized co-firing biomass (mostly wood pellets) in coal power plants. The second major contributor to rising biomass-CO2 emissions is the waste sector. The increase may be due to Dutch waste management policy shifting focus from waste disposal to energy recovery (see section on CH4). While emissions from the waste sector peaked in 2016 and have since stabilized, biomass CO2 emissions in the energy sector continue to rise.

Different forces contribute to changes in non-biomass CO2 emissions

Non-biomass CO2 emissions have fallen below 1990 levels only since 2020 (see figure 5). These reductions are partly due to the Covid-19 pandemic, which significantly slowed economic activities worldwide. Sectors such as aviation, transportation and trade faced major disruptions in both output and emissions. The reduction in GHG emissions observed during the Covid crisis is seen as a temporary effect (cyclical), rather than a lasting change (trend) in the overall pattern of emissions. When re-estimating the trend relationship between non-biomass CO2 emissions and GDP, we find a negative trend elasticity, indicating absolute decoupling. However, this relationship is less statistically significant compared to the elasticities found for methane and nitrous oxide (see appendix 3).

As there are many sources of non-biomass CO2 emissions, we focus on three biggest emission sources: manufacturing, energy, and transportation (see figure 7). Below, we discuss these sectors in more detail.

The manufacturing sector

Within manufacturing, the chemical industry, basic metal industry, and petroleum industry are the top three emitters of non-biomass CO2 emissions. Figure 8 shows the changes in non-biomass CO2 emissions for these three sectors.

CO2 emissions from these three sectors are concentrated in a small number of companies. According to the latest figures from the Dutch Emissions Authority (NEa) (in Dutch), of all Dutch installations under the European Emission Trading System (EU ETS), the top 12 manufacturing companies account for over 70% of total manufacturing CO2 emissions. Among these, seven chemical companies are responsible for over 90% of the chemical industry’s CO2 emissions, four refinery companies account for 60% of the oil industry’s CO2 emissions, and one company is responsible for 90% of the basic metal industry’s CO2 emissions. Therefore, short-term fluctuations in CO2 emissions from the manufacturing sector heavily depend on the dynamics of these companies. This is recognized by the ministry of Economic Affairs, which has implemented a framework called Maatwerkafspraken (in Dutch) to support top industrial emitters in reducing their emissions.

From 1990 to 2022, more efficient use of energy contributed to a reduction in CO2 emissions in manufacturing, as reflected in the declining energy intensity in the sector. Meanwhile, higher energy prices have led to declining investments and less production activities in some companies, which also resulted in lower CO2 emissions in these sectors. Over the past twenty years, some companies in these sectors have scaled back their investments in production facilities due to the relatively high cost of energy in the Netherlands. Furthermore, in 2022, some manufacturing companies (including Chemelot, Nyrstar and Yara (in Dutch)) partially or completely halted their activities in the Netherlands due to high energy prices. The most recent CO2 emissions and efficiency data from NEa (in Dutch) also reveals that the drop in CO2 emissions in manufacturing after 2021 is mainly due to lower industrial production and only partly the result of improved and cleaner production methods.

The energy sector

Emissions from the energy sector only fell below the 1990 levels starting in 2020 (see figure 7). However, there has been an accelerated decline in non-biomass CO2 emissions due to the rapid and substantial adoption of non-biomass renewable energy sources, such as solar and wind, in recent years (see figure 9). While non-biomass CO2 emissions have decreased, biomass CO2 emissions have actually increased. The increase in CO2 emissions from biomass suggests that just because a fuel is renewable, it doesn’t mean it will automatically reduce CO2 emissions. This highlights the complexity of trying to achieve multiple environmental goals with a single policy, as the renewability of biomass doesn’t necessarily align with lower emissions.

The transport sector

A growing economy increases the volume of goods (in Dutch) transported by road, sea, and air. After 2007, the emissions from the transport sector have shown a declining trend, primarily due to improvements in vehicle fuel efficiency. The EU has progressively introduced stricter emissions standards for road vehicles, such as the Euro 5 standards in 2009 and the Euro 6 standards in 2014. In the Netherlands, vehicles must meet these standards to access so-called “milieuzones” (environmental zones). Additionally, fiscal incentives (in Dutch) for purchasing fuel-efficient cars have increased, leading to a rise in the share of electric vehicles. However, the pace at which this trend can continue in the transport sector remains unclear due to various policy changes.

Reduction in methane (CH4) is primarily driven by change in landfill regulations

The trend elasticity estimates for CH4 are significantly negative and much larger (-1.19) compared to the overall GHG trend elasticity. This suggests an absolute decoupling of CH4 emissions from GDP growth. However, these results can be misleading, as the reduction in CH4 emissions is mainly due to changes in landfill practices rather than economic activity (see figure 10).[3] Landfilling is a disposal method for solid waste, where waste decomposes in the absence of oxygen, producing and releasing CH4. Since 1990, changes in Dutch landfill regulations have altered the types of materials being sent to landfills, resulting in lower CH4 emissions. This underscores the importance of policy intervention in reducing emissions. However, changes in waste treatment may have led to higher CO2 emissions, as waste is diverted from landfilling to incineration for energy recovery.

[3] According to SEEA guidelines, emissions from landfills don’t originate directly from a production process but from “accumulation,” so they should be recorded separately.

Next to landfills, the agriculture sector has been the largest contributor toCH4 emissions reduction among all economic sectors from 1990 to 2022. In 2022, agriculture accounted for 76% of all Dutch CH4 emissions. Within agriculture, CH4 emissions were lowest 2004 and 2005 compared to 1990, which aligns with the lowest number of cattle in the Netherlands around 2005. Some of the progress in CH4 emissions reduction has since been lost. The removal of the milk production quota by the European Commission, led to an increase in the number of animals (in Dutch), but this trend was reversed in 2018 with the introduction of phosphate rights (in Dutch). Both policy changes had an impact on CH4 emissions from Dutch agriculture, contributing to an increase in CH4 emissions between 2014 and 2017, followed by a reduction from 2017 to 2022.

Lower nitrous oxide (N2O) emissions also coincide with policy changes

The two main sources of N2O in the Netherlands are the agriculture and manufacturing sectors. The sharp reduction in N2O emissions is mainly driven by lower emissions in the chemicals sector and agriculture (see figure 11). Agriculture led the reduction in N2O emissions until the early 2000s and has since stabilized. Notably, in 2008, a sharp decrease in N2O emissions by the chemical industry led to a significant decline in overall N2O emissions compared to 2007. This decrease coincides with changes in nitric acid production (in Dutch) as a result of this sector coming under the European Emission Trading System (ETS).

The analysis has so far been focused on the observed historic relationship between Dutch emissions and GDP. As stated in the introduction, this relationship (and its future trajectory) lies at the heart of European policies aimed at achieving the Paris Agreement targets. One way to assess our progress toward these targets is by using the Kaya Identity. This accounting identity identifies population, GDP per capita, energy intensity, and carbon intensity as main drivers of CO2 emissions (see appendix 2).

In figure 12, the horizontal axis represents levels for the four drivers in 1990. A positive bar for a given driver in any given year indicates that its level in that year is higher than its 1990 level, while a negative bar indicates that it is lower. The magnitude of the bar represents the growth rate relative to its 1990 level. For example, in 2022, Dutch GDP per capita was 63% higher, the population was 18% higher, the energy intensity of GDP was 50% lower, and the carbon intensity of energy was 2% lower than their respective 1990 levels. Data up to 2022 is based on historical observations from CBS data, while numbers for 2023 and beyond are projections based on the sources explained below.

Challenge in meeting Dutch 2030 emissions targets

The Netherlands aims to reduce GHG emissions by at least 55% by 2030 compared to 1990 levels. This translates into an annual compound growth rate of -8.7% in carbon dioxide equivalent (CO2e) emissions. We have combined this with the CBS population forecast for 2030 (18.35m) and GDP growth forecasts from RaboResearch (in Dutch). The energy intensity forecast reflects the government’s National Energy and Climate Plan. Using these inputs, the Kaya Identity allows us to calculate the decline in carbon intensity needed to meet the Paris climate agreement targets.

The most recent data (for 2022) shows that the carbon intensity of energy is 2% below the 1990 levels. It needs to be almost 42% lower by 2030 for the Netherlands to meet its stated 55% emissions reduction target. We cannot estimate the effects of planned policies (such as changes in the European ETS) on the energy and carbon intensity rates. However, the PBL has investigated impacts of planned policies and finds that the 2030 target is out of reach(in Dutch).

As non-CO2 emissions have been the most important drivers behind lower GHG emissions, there may be limited potential for further reductions. Additionally, CO2 emissions remain the largest share of GHG emissions. Both factors stress the sense of urgency to reduce CO2 emissions even faster.

We benefited from discussions with many colleagues and specifically thank Ester Barendregt, Harry Smit, Barend Bekamp, Yorick Cramer, and Erik Lokhorst for their valuable input and feedback on this work.

Cohen. G., Jalles, J., Loungani, P. & Marto, R. (2018). The long-run decoupling of emissions and output: Evidence from the largest emitters, Energy Policy, 118, 58-68.

Davis, S., & Caldeira, K. (2010). Consumption-based accounting of CO2 emissions, Proceedings of the National Academy of Sciences, 107(12), 5687-92.

Doda, B. (2014). Evidence on business cycles and CO2 emissions, Journal of Macroeconomics, 40, 214-227.

Heutel, G. (2012). How should environmental policy respond to business cycles? Optimal policy under persistent productivity shocks, Review of Economic Dynamics, 15(2), 244-264.

Kaika, D., & Zervas, E. (2013). The Environmental Kuznets Curve (EKC) theory—Part A: Concept, causes and the CO2 emissions case, Energy Policy, 62, 1392-1402.

Kaya, Y. , & Yokobori, K. (1997). Environment, energy, and economy: Strategies for sustainability, United Nations University Press.

Kriström, B., & Lundgren, T. (2005). Swedish CO2-emissions 1900–2010: An exploratory note, Energy Policy 33(9), 1223-1230.

Liddle, B., & Messinis, G. (2018). Revisiting carbon Kuznets curves with endogenous breaks modeling: evidence of decoupling and saturation (but few inverted-Us) for individual OECD countries, Empirical Economics 54, 783–798.

Narayan, P., & Narayan, S. (2010). Carbon dioxide emissions and economic growth: Panel data evidence from developing countries, Energy Policy, 38(1), 661-666.

Pao, H., & Tsai, C. (2010). CO2 emissions, energy consumption and economic growth in BRIC countries, Energy Policy, 38(12), 7850-7860.

Peters, G., Minx, J., Weber, C., & Edenhofer, O. (2011). Growth in emission transfers via international trade from 1990 to 2008, Proceedings of the national academy of sciences, 108(21), 8903-8908.

Stern, D. I. (2004). The rise and fall of the environmental Kuznets curve, World Development, 32(8), 1419-1439.

Summary of key literature on the relationship between emissions and economic activity

Ever since the industrial revolution, the unprecedented global economic growth has been primarily powered by fossil fuels. As a result, GHG emissions have historically gone hand in hand with economic growth. However, the International Energy Agency suggests that this relationship weakened over time for most countries. As economies develop, environmental pressures initially increase but then decrease – a phenomenon known as Environmental Kuznets Curve (EKC). Academic literature also shows that this relationship has evolved over time. For example, Kriström and Lundgren (2005) found different trends in emissions in Sweden over long periods (1900-1999) compared to shorter ones (1970-1999). Similarly, Cohen et al. (2018) discovered that for 20 countries, the trend elasticity of emissions to GDP post-1990 is much smaller than from 1946 onward.

Stern (2004) and Kaika and Zervas (2013) review extensive literature on how economic development and associated negative environmental impacts (such as emissions, pollution) co-evolve over the development stages of an economy. While early results supported the EKC, more recent studies question the inverse U-shaped relationship between emissions and economic activity. This indicates that negative environmental impacts do not always increase and then decrease as an economy develops. Therefore, environmental problems don’t resolve themselves and require explicit attention. The results vary depending on choice of data, country, time horizons, and econometric methods.

The importance of distinguishing between cyclical and trend relationships

Academic literature analyzes the emission-output relationship by distinguishing business cycles (short run) and trends (long run).

The cyclical relationship between GHG emissions and GDP is based on the notion that emissions respond to GDP fluctuations over business cycles. Heutel (2012) explores the higher volatility and pro-cyclicality of emissions in the US. Doda (2014) extends Heutel’s analysis (2012) to a panel of countries, findingthat emissions are procyclical over short-term business cycles, meaning emissions tend to be above their trend during economic expansions and below during recessions.

There is also a long-term trend relationship between emissions and GDP, focusing on the long-run co-movements in emissions and output. For example, Narayan and Narayan (2010) use a panel cointegration model to estimate short-run and long-run elasticities for 43 developing countries. They found long-run elasticities to be smaller than short-run elasticities for the Middle Eastern and South Asian panels. Pao and Tsai (2010) estimate long-run elasticities for a panel of BRIC countries (Brazil, China, India, and Russia), and found that output exhibits the inverted U-shape pattern associated with the EKC hypothesis.

It is difficult to assess long-term co-movements of GDP and emissions if they are masked by cyclical fluctuations. Cohen et al. (2018) use a simple trend/cycle decomposition to analyze the world’s top 20 emitters. They found that once the cyclical relationship is accounted for, the trends show evidence of decoupling between emissions and economic growth in richer countries, particularly in Europe, but not yet in emerging markets.

GHG emissions embedded in international trade

Existing literature highlights the significant role of international trade in the transition to a low-carbon path. The World Trade Organization suggested in 2021 that developed economies tend to be net importers of greenhouse gas emissions, while developing countries tend to be net exporters. Davis and Caldeira (2010) found that, in 2004, nearly a quarter of global CO2 emissions were embodied in exports from China and other emerging markets to more advanced economies. Peters et al. (2011) documented that net emissions associated with trade flows from developing to developed countries increased fourfold between 2000 and 2008.

For the Netherlands, GHG footprint statistics (in Dutch) from CBS show that between 2008 and 2021, the emissions trade balance – the emissions abroad associated with goods and services produced for Dutch consumption minus the emissions in the Netherlands associated with goods and services produced for consumption in other countries – was positive throughout the period. This positive emissions trade balance, combined with the Dutch emissions associated with economic activity physically based in the Netherlands (production-based emissions), results in a higher total value of emissions attributable to the Netherlands. This higher value is often used to calculate consumption-based emissions in the carbon footprinting literature.

When accounting for emissions from international trade, Cohen et al. (2018) conclude that the decoupling of GHG emissions from economic growth for richer nations becomes weaker. This underscores the sensitivity of the relationship between emissions and GDP to the use of a particular emissions metric (production-based or consumption-based).

Explanation of choices made in the study setup

A narrow definition of decoupling

We adopt a narrow definition of “decoupling” to examine, solely the relationship between GHG emissions and economic activities. This approach leaves out the broader and more complex discussion about decoupling economic growth from environmental degradation (see this report from European Environmental Bureau for a comprehensive definition).

We focus on one individual country – the Netherlands – instead of a group of countries

Pooling information over time and across countries (panel econometric methods) is efficient for exploiting more information, but it cannot reflect the unique development pattern of each individual country. Even within European countries, the emission-output relationship can differ significantly (see Liddle and Messinis, 2016). We have detailed data on the Dutch economy and emissions from CBS, which allows a more precise analysis.

Our focus period runs from 1990 to 2022

The year 1990 is highly relevant for GHG emissions. The nationally determined contributions submitted by the European Council under the Paris Agreement set a target of at least a 55% reduction in GHG emissions by 2030, compared to 1990 levels. This target is also binding for the Netherlands and has been endorsed as a policy goal by the Dutch government. All emissions and GDP figures in this report use 1990 as the reference year. The end year, 2022, is determined by the availability of data.

Our report follows the literature on cycle/trend decomposition and uses the trend component to estimate the emission-output elasticity, referred to as trend elasticity.

Literature suggests that it is important to analyze the emission-output relationship by distinguishing business cycles (short term) from trends (long term) (see the literature discussion above ). We follow this approach by removing the cyclical fluctuations in both GHG emissions and GDP before assessing the decoupling as measured by the trend elasticity. The trend elasticity is the percent change in emissions for a one percent change in output over the long term. We use this elasticity as a measure to assess the direction and strength of long-term relationships between emissions and GDP.

Appendix 2: provides a detailed explanation on the methodology.

An elasticity below unity (1.0) points to relative decoupling, where 1% economic growth results in a less than a 1% rise in GHG emissions. An elasticity below 0 indicates absolute decoupling, where 1% economic growth is accompanied by a reduction in GHG emissions. Appendix 3 presents the estimates of trend elasticities, which are referred to in the subsequent sections. We do not estimate or focus on elasticity based on the cyclical component.

Finally, we use production-based emissions to measure GHG emissions by Dutch economic activity (in Dutch), rather than consumption-based emissions. The “footprint” shows which emissions are related to the consumption of goods and services by Dutch residents (see appendix 4). CBS data onthe Dutch GHG footprint (in Dutch) (available since 2008, with some gap years) indicates that the Netherlands has emitted more GHGs from its consumption than from its production. This is because the goods it imports are relatively more emission-intensive than the goods it exports. While literature suggests that accounting for consumption-based emissions can affect the relationship between GHG emissions and economic growth (see the literature discussion above ), we don’t include a footprint analysis in our current report. Measures for consumption-based emissions vary across different databases (such as Global Carbon Budget and the Eora multi-region input-output (MRIO) database) as they use different models to estimate these emissions, making it hard to choose. Furthermore, it is challenging to link consumption-based emissions to specific economic activities in different countries or regions. In contrast, production-based emissions are easier to link to economic activities and can better explain the underlying factors of development of emissions. That said, footprint analysis can provide better insight into the Dutch contribution to global emissions.

Detailed explanation of the HP filter and the trend elasticity

To analyze the long-run movement of GHG emissions and economic output, we distinguish between trends and cycles in both emissions and output. We have used the Hodrick-Prescott filter to extract the cyclical and trend components. This filter minimizes the following function:

Where , is the trend component and λ is the smoothing parameter (set at 100, which is common practice when employing annual data). The difference between x_t and the trend component is , the cyclical component.

The following two figures display Dutch GHG and GDP year-on-year growth in aggregate versus decomposed trend components.

After decomposition, we estimate the trend emission-output elasticity using the following specification:

Where is the trend of the log of emissions, and is the trend of the log of real output. is the trend elasticity, which is the focus of our report. It measures the percentage change in emissions resulting from a one percent change in output. If it is positive but less than one, emissions increase less rapidly than output, implying relative decoupling. If it is negative, it indicates absolute decoupling between emissions and GDP.

Description of the Kaya Identity and its application

The Kaya Identity is an accounting identity developed by the Japanese energy economist Yoichi Kaya at an IPCC seminar in 1989 (see Kaya and Yokobori, 1997). This identity is useful for attributing CO2 emissions to four underlying driving factors: population size, per capita GDP, energy intensity of GDP, and carbon intensity of energy supply. The accounting identity can be stated as follows:

where

CO2_t = CO2 emissions rate (in million kg) in year t.

POP_t = population (number of people) in year t.

GDP_t = gross domestic product (EUR, constant prices of 2015) in year t.

TES_t = total energy supply (in Exajoule, EJ) in year t.

In this identity:

GDP/POP denotes GDP per capita.

TES/GDP denotes energy intensity of GDP.

CO2/TES denotes CO2 emissions per unit of TES, which is the carbon intensity of energy supply.

A log transformation on both sides of the identity states that the change in CO2 emissions is the sum of changes in population, GDP per capita, energy intensity of GDP, and carbon intensity of energy.

Comparison of different frameworks for calculating GHG emissions

There are various frameworks for calculating GHG emissions. One framework is environmental accounts, which focuses on GHG emissions from production activities within the Dutch economy, excluding emissions from the Land Use sector. Another important framework is IPCC, which includes emissions from human activities within a specific country.

The differences between GHG emissions measured by Environmental Accounts and the IPCC are summarized as follows (based on the CBS article, in Dutch):

Combustion of biomass: The IPCC framework does not include the combustion of biomass because it is considered short-cyclic. The IPCC assumes that the CO2 released during biomass combustion is recaptured in biomass, thus not contributing to an increase in the CO2 concentration in the atmosphere. This includes the co-firing of biomass in power stations, as well as the consumption of biogas and transport fuels such as biodiesel.

International aviation and international shipping: In the Environmental Accounts framework, emissions from international aviation and international shipping are counted as economic activities within the calculation framework of the Environmental Accounts. This concerns the emissions of Dutch airlines and shipping companies, regardless of where they refuel. These emissions are not included in the national IPCC emission totals but must be reported in a separate memo.

Territorial versus resident-based emissions: The IPCC framework is based on emissions within Dutch territory, while the Environmental Accounts framework is based on activities by Dutch residents. For example, the IPCC calculates road traffic emissions based on motor fuels refuelled in the Netherlands, regardless of the origin of the vehicle. In contrast, the environmental accounts framework considers the kilometers driven by Dutch residents, regardless of where the driving occurs.

Discussion on production-based versus consumption-based emissions

The footprint shows the emissions related to the consumption of goods and services by Dutch residents. This includes the emissions generated abroad for the production of goods and services that are ultimately consumed in the Netherlands. Conversely, foreign consumption with production in the Netherlands does not count toward the Dutch footprint.

When examining decoupling within an economy, it’s preferable to focus not solely on emissions from economic activities within the borders of an economy but also on emissions associated with the total global supply chain. More specifically, certain emission-intensive sectors in the Netherlands may have outsourced the emission-intensive part in the value chain to emerging markets. This may have lowered the emission intensity of domestic production relative to the domestic value-added development in those sectors. However, through the import of emission-intensive intermediate goods, the global emissions associated with the production of final goods in these specific industries may not have decreased.

In this context, and as mentioned earlier, focusing on consumption-based emissions in relation to economic activity helps address the issue of accurately measuring emissions. By adjusting the emissions data to account for trade, we include the emissions from both imported and exported goods. This gives a more complete picture of the emissions associated with the total supply chain.