11.1: The Emissions Challenge

There are few machines that have had a bigger impact on human development than the steam engine. While the first steam engines go back to the first century AD, it was only in the 1700s that they were used in a production context. While at first they were used to pump water out of mines, by the late 1700s the Boulton-Watt engine was able to turn steam into rotative motion. This game-changing invention led to a massive increase in the demand for steam, which was most easily generated by combusting fossil fuels (for example, coal, gas, oil) and renewables (for example, biomass, wood). This invention was instrumental in kicking off the Industrial Revolution, which led to a massive expansion in the demand for fossil fuels. Further uses for fossil fuels came from the smelting of steel; the heating of homes, factories, and places of employment; and the arrival of the internal combustion engine, which enabled today’s main modes of transportation.

Where do the emissions come from?

The inventions powered by fossil fuels led to the explosive growth in emissions of greenhouse gases that continues to the present day. Figure 11.1.1 displays the growth of CO₂ emissions from fossil fuel combustion since the 1700s. The world essentially went from negligible emissions to 35 billion tons of CO₂ in 2013. Most of this growth stems from the combustion of coal (solids), oil and its derivative products (liquids), and natural gas. There is a significant single use—the production of cement, which shows up in these graphs because the chemical process used to make cement generates CO₂ directly. Finally, when oil and gas are produced, some gas escapes and is not captured, for cost-effectiveness reasons, and is flared (burned) off.

Figure 11.1.1 Carbon dioxide emissions by fuel source globally from 1750 until 2014. Because of the long lifetime of CO2 in the atmosphere, it is not the level of emissions in any given year, but the area under the curve that matters. From Ritchie and Roser 2019.

Figure 11.1.2 Global emissions by world region. For the latter part of the twentieth century, the United States was the leading emitter, but it has recently been overtaken by the People’s Republic of China. From Ritchie and Roser 2019.

Instead of looking at emissions trends by fuels, it is instructive to break down emissions by region. Figure 11.1.2 replicates Figure 11.1.1 but indicates the source region of the emissions. Three things stand out from this figure. First, most of the emissions growth until the end of the twentieth century came from high-income countries, with the European Union (EU) and the United States being responsible for the lion’s share of emissions. What we see in the twenty-first century, however, is that the most significant growth in emissions has come from lower and middle-income countries. Most notably, China became the biggest emitter (in total, not per capita terms) early in the twenty-first century, and we are observing significant growth in India’s emissions as well as those of the rest of Asia. Emissions in Europe and the United States have leveled off or stopped their significant growth path—for now.

What is driving emissions?

Figure 11.1.3 Per capita CO2 emissions plotted against per capita GDP (the value of goods and services produced per person) across countries. The relationship is plotted on a log scale, which gives the illusion of a linear relationship between the two variables, when in fact the relationship looks more like the right part of a parabola. From Ritchie and Roser 2019.

The question as to what is driving this growth is not as mysterious as one thinks. It has been formalized in what is called the Kaya identity. The identity states that CO2 emissions can be expressed as a product of population, GDP per capita, energy intensity per unit of GDP, and carbon intensity (carbon per unit of energy consumed). It hence suggests that more populous, richer, more energy-intensive countries with a more carbon-intensive energy sector will have higher emissions. This is not rocket science and has been confirmed across countries through a variety of research efforts. As this chapter focuses on the economic aspects of climate change, taking a closer look at the income-emissions relationship is instructive. As China has a much larger population than the US, for example, a meaningful comparison here is to look at the relationship between per capita GDP in a common currency and emissions per capita. This gives us a snapshot of how average income and emissions correlate. Figure 11.1.3 plots this for the year 2016.

I find this figure enlightening. What we see here, albeit on a log-transformed scale, is that there is a strong positive relationship between per capita income and per capita emissions. Some sub-Saharan economies, such as Burundi, which are among the poorest in the world, have GDPs below $1,000 per person per year and per capita emissions below 0.1 tons per year. The world’s richest countries—the United States, for example—have GDPs around $60,000 per person per year and have per capita emissions north of 10 tons per year. Yes. Almost 100 times the emissions per person of the world’s poorest individuals.

While this graph is enlightening in the cross section, it is not surprising. Richer countries consume (and produce) more things and hence have higher emissions. This snapshot, by definition, ignores time. What if some or most of the bigger countries in the bottom left quadrant of Figure 11.1.3 migrate to the top right quadrant? Is this a bad thing? That depends on what you tend to worry about. The Kaya identity suggests that if you get richer, you emit more, which is intuitive. Also, as your economy becomes more carbon intensive—as many economies transitioning from an agrarian economy to an industrialized economy have—you emit more carbon. At least that is what the historical record suggests. One hope is that the amount of energy needed to produce a unit of output will go down as technology becomes more efficient. We have observed this in many industries. One example is lightbulbs. Old incandescent lightbulbs used much more electricity and provided less light than new LED lightbulbs do.

One thing to wonder about, at least according to the Kaya identity, is whether rising incomes are a bad thing. If you were a development economist, you would be excited about rising incomes! Battling poverty is one of the main goals of development economists and practitioners. In fact, the first Millennium Development Goal (MDG) stated by the United Nations is to “eradicate extreme poverty and hunger.” This has historically come at an environmental cost—a rise in greenhouse gas emissions. If you read down the list of the MDGs, the seventh MDG is to “ensure environmental sustainability.” What this would suggest is that we are seeking a global transformation of human well-being, by eliminating suffering, all the while ensuring environmental sustainability. What this means is that we would like to move all countries to the right in Figure 11.1.3, while ensuring that countries currently low on the y-axis remain there and, at the same time, bring the countries in the top right quadrant down into the bottom right quadrant by reducing their emissions but preserving their wealth. This is clearly not a small task and will require a Herculean effort by politicians and researchers across the globe.

The global nature of current and future emissions

While it is easy to start thinking about what is happening in individual countries after seeing Figure 11.1.3, it is important to remember that greenhouse gases are largely global pollutants. A ton of CO2 is roughly what you would emit by driving a Ford Mustang 5.0 (one cool car, especially the convertible) from San Francisco to Chicago. It does not matter whether that ton of CO2 is emitted in California or China—it causes the same amount of damage. Hence the global climate responds to global emissions, regardless of where they stem from. There are numerous efforts under way to draw different scenarios of global emissions for the next 100 years. As you can imagine, predicting what will happen hundreds of years from now is extremely challenging. Imagine trying to predict today’s economy and technology as a scientist living in the 1700s! These emissions scenarios are consistent with different versions of the Kaya identity. Figure 11.1.4 displays a set of future emissions scenarios.

Figure 11.1.4 Possible scenarios of greenhouse gas emissions for the next 100 years. You can see that the scenario without climate policies is projected to cause significant warming, while the dark blue path is an extremely ambitious emissions scenario limiting warming to 1.5°C. From Ritchie and Roser 2019.

The top (peach-colored) band in Figure 11.1.4 displays a future where no climate policies are implemented—the status quo; the green trajectory shows a future where currently implemented climate policies are executed and adhered to. The purple band displays a future where all countries achieve their current targets/pledges set within the Paris Agreement. The red pathway corresponds to an aspirational goal of limiting emissions to a level where the global temperature only rises by 2°C above preindustrial levels. This would limit damages to avoid some extreme and worrisome effects. Finally, the blue trajectory is for an aspirational goal of limiting warming to 1.5°C. This scenario would require very urgent and rapid reduction in global greenhouse gas emissions. Figure 11.1.4 is just one example of a set of emissions scenarios produced in a very active academic literature. Of course, you and I could probably come up with alternate futures, which would look very different.

The emissions scenarios are used by physical climate scientists as inputs into climate models (general circulation models, sometimes called global climate models [GCMs]). These models, as discussed in Chapter 1, translate changes in greenhouse gas emissions into impacts on physical dimensions of climate. These include, but are not limited to, surface temperature, precipitation, humidity, wind speeds, cloud formation, and sea level rise. These models hence produce different futures of planet Earth depending on what emissions pathway we follow. How big changes in the climate system will be depends critically on the emission path we will follow.

Figure 11.1.5 Projected effects of climate change on temperature, precipitation, sea ice, and ocean acidity under lower (left) and higher (right) emissions scenarios. From IPCC 2013.

The Intergovernmental Panel on Climate Change (IPCC) generates the “official” climate scenarios, which are a synthesis of the mostly peer-reviewed scientific literature to date. Thousands of scientists review tens of thousands of published papers to synthesize our collective understanding of the current and future state of the climate system. Figure 11.1.5 displays the projections from their Fifth Assessment Report. There are different scenarios of climate change, called Representative Concentration Pathways (RCPs), which are simply worlds with different degrees of greenhouse gas concentrations. The graphs on the left display temperature change, precipitation change, and Arctic sea ice under RCP 2.6, which is consistent with warming of roughly 1°C over temperatures experienced in the late 1900s. Under this scenario the chance of exceeding 2°C is less than 33%. The graphs on the right show the same indicators (temperature, precipitation, and sea ice) under the high RCP 8.5 emissions scenario, for which temperatures are thought to continue increasing and reach about 4°C higher than late-twentieth-century levels (the likely range of outcomes for 2100 is 3°C to 5.5°C higher).*

The figure shows a number of interesting things. First off, the impacts on temperature are much more dire for the higher-emissions scenario (which is not that surprising) and for regions closer to the poles. Second, most climate models predict very similar increases in warming and distribution across space for temperature. The second row in Figure 11.1.5 indicates changes in precipitation. Two things stand out here. First, the high-emissions scenario has very different patterns of drying in different areas, and those with the largest decrease in precipitation include some of the major crop-producing regions in the world. Second, the model agreement is much lower, with different models predicting very different precipitation changes, making this one of the most important uncertainties in the climate literature. Especially if you are an agricultural economist like me! Crops need water to grow, and what and where to plant is one of the key decisions a farmer has to make. The more uncertainty there is in what your rainfall patterns look like, the harder it gets to make the best planting decisions. The third row shows Arctic sea ice under the two scenarios. It shows that under both scenarios, sea ice shrinks. Most notably it disappears under the RCP 8.5 scenario.

A pet peeve of mine is that funding agencies across the world have spent billions of dollars studying the physical aspects of climate change by collecting important data and supporting very complex computational models to study the future of the climate system. While this is clearly extremely important, much less attention (and, importantly, research funding) has been directed at studying the impacts of climate change on human and natural systems and translating these into monetary terms. Lack of funding notwithstanding, there has been a sparsely funded explosion in this literature recently, which led to the development of methods and insights that were not available as far back as 10 years. Much of this revolution in economics has been fueled by the availability of large data sets on the economy and detailed imagery data collected via high-resolution satellites.

*Note that we keep on using °C, which is the unit used in most of the literature (water freezes at 0°C and boils at 100°C). If you like Fahrenheit better, multiply the degree Celsius figure by 9, then divide by 5, and add 32 to what you get.