Chapter 9
As was mentioned in Chapter 8, the essential role that both the socioeconomic human factor and the science of the physical environment play in modeling future climate changes makes the issue very complex. A method was proposed in the 1970s to simplify the issue by formally presenting the increase in GHGs as a product of various factors so that dimensions of the various terms will cancel out. This kind of analysis is referred to as dimensional analysis. In the context of environmental issues, this methodology is known as the IPAT identity,1– 3 where I stands for impact, P for population, A for affluence, and T for technology.
For emission of CO2 the identity can take the following form:
. [9.1]
F ossil f uels
CO2
F ossil f uels
Energ y
Energ y
GDP
125
GDP
CO= Popula tion
2
⎞
⎛
⎞
⎛
⎞
⎛
⎞
⎛
⎟⎠
⎜⎝
⎟⎠
⎜⎝
⎟⎠
⎜⎝
⎟⎠
⎜⎝
GDP/population (or as it is more often expressed as GDP/capita) is the measure of affl uence. This term was already discussed in Chapter 2 in the context of the linear relationship with energy use. It will be further discussed here and in more detail in Chapter 12. The rest of the terms are the technology part of the identity.
In this chapter I will discuss the first three terms of this equation, which is the foundation of the present political, global discussion, often referred to as the Kyoto process, to wrestle with these issues. This process will be covered in Chapter 13. The discussion in this chapter is based on data from United Nations (UN) and World Bank databases—data for which are supplied mostly by individual countries. These data often carry agendas beyond an academic interest and are often criticized for that reason. The advantage of these data is that they are transparent and are freely available through the Internet. The numbers are more or less compatible with other available databases and any serious challenge to these data can be immediately explored.
The IPAT equation is convenient for discussion because it allows us to diff erentiate the various contributors to climate change. It is not necessarily correct—one can achieve dimensional agreements through other devices. However, one can examine the validity through the available data and conclude that, although simplified, it is essentially valid. Our discussion will be based on this equation. In this chapter I will discuss the first three terms on the right side of the equation and in the next chapter I will examine the last two terms.
Figure 9.1 shows the UN estimate for changes in the world’s population over the next 50 years. The world’s present population is about 6 billion, and it is estimated that in about 50 years, the population will increase by about 50% to about 9 billion people. One can see from the figure that the rate of global population increase (the slope of the curve) decreases with time. It is estimated that this decrease will continue until we reach a stable population of about 10– 11 billion people toward the end of the 21st century. Similar to the projections in Chapter 8, these population change estimates are based on models of the socioeconomic developments in the near future (within our “now ” period discussed in Chapter 2). The estimate in Figure 9.1 is the UN’s median estimate— one can find high estimates and low estimates based on somewhat different scenarios— just like climate modeling in Chapter 8. Since these estimates are available for the last 50 years, one can compare their validity with the actual population growth. The median estimate is doing well by this measure, and is the one that I am using here. Th e projected population growth is not geographically uniform. This is shown in Figure 9.2.
Figure 9.2 shows the projected continental distribution of population growth. Asia dominates the growth picture and is projected to increase its share of the world’s population. Perhaps
Figure 9.1. UN median estimate of the world population Source: UN Department of Economic and Social Aff airs (2004).4
Figure 9.2. Projected continental distribution of estimated population growth Source: UN Department of Economic and Social Aff airs (2004).4
more important than the projected continental distribution is the distribution of wealth and its correlation with demographic distribution.
Figure 9.3 shows the population growth rate for 10 countries that together account for more than 50% of the world’s population as a function of their GDP per capita. There are serious complications one encounters when using GDP per capita to measure a country’s wealth. Most important among these is that many factors that directly influence GDP per capita have nothing to do with the wealth of a country. Chief among these factors are devaluation of currency and inflation in the country that serves as a reference. To account for some of these issues, the measure in Figure 9.3 is normalized to US$ in one particular year (1995). Usually more adjustments are needed, and some will be reflected in data presented in this and subsequent chapters. This topic will be discussed further in Chapter 12.
In Figure 9.3 I compare the population growth rates for 1975 and 2003. Inspection of this figure immediately brings to our attention the most important trends that led to the UN projections together with some inadequacies of my presentation here. Let us start with the
Figure 9.3. Population growth as a function of GDP per capita for the following countries (in increasing GDP per capita for 2003): India, China, the Philippines, Egypt, Russia, Mexico, Brazil, France, the United States, and Germany
Source: World Bank.5
latter— a critical reader will observe that the data for 10 countries are shown in the 2003 population growth data while only 9 are shown in the 1975 data. In addition, the population growth of one poor country in the 2003 data drops way below the general trend. Both issues revolve around Russia. I could have eliminated both issues by not including Russia in the 2003 data, but I chose not to do that. The reason for Russia’s behavior is clear— in 1975 Russia was part of the Soviet Union, and the World Bank database did not keep separate statistics for Russia. Thus its separate GDP was not available, and the drop in the population growth rate reflects the upheaval. I chose to include Russia mainly to emphasize two points: First, aside from esthetic considerations, we do not sacrifi ce any information by including Russia in the 2003 data— we do not take any averages to be distorted by Russia’s abnormal behavior, and we clearly mark the value of each country. Second, to get general trends we should be able to filter out stark deviations from the trend after satisfactorily accounting for the deviations specific to each country that do not affect the trend. Two general trends can be easily observed: in both years the population growth of the richer countries is lower than for the poorer countries, and there is a marked decrease in population growth for the rich and poor countries from 1975 to 2003. The poorer countries are experiencing a larger decrease in population growth than the richer countries (because they started from a much higher base). Th ese conclusions are not based on the data in Figure 9.3; it is always dangerous to stipulate a trend based on an extrapolation between two data points. The World Bank and similar data sources provide a continuous set of data (subject to availability) from around 1950, yearly, for all countries. Th e trends can withstand the scrutiny of adding more countries at more frequent time intervals. The predicted saturation of the world’s population (a population growth of zero) can already be seen in the developed world.
W hat are the reasons for the observed global decline in population growth crucial to the predictions of population impact on environmental changes? The underlying reasons are the topic of intense research efforts— which usually means that the issue is important and not well understood. It might also mean that we need some historical perspective in order to sort out important causes from less important ones. In all the main scenarios people use to try to explain the trend, women’s increased participation in decision making, education, and professional and societal aspirations play a central role. However, many questions remain. For instance, GDP per capita is an average number that reflects the economic well- being of a country. Fertility rates in most countries, with some very notable exceptions such as China, are results of individual decisions. Some of these decisions are the result of individuals acquiring control over reproductive processes, and some are not. The declining population growth can also be seen in theocratic countries such as Iran and Saudi Arabia, in which major obstacles still remain for women’s equality.
W hatever the causes, the trend is clear— a major decrease in population growth will gradually stop the global population increase in stages. First, the population of the developed, rich countries will reach zero growth, a process that in many developed countries is already on its way. This will be followed by population saturation in developing countries. The result of this differential rate is that toward the end of the century, when the population is estimated to reach 10– 11 billion, 90% of this population will reside in countries now classified as developing.
Can the world be divided into rich countries and poor countries, and what are the dividing lines?
Currently (2003), the UN is working with the following classification: high- income countries are the 56 countries with GDP per capita greater than US$9000— they include most European countries, the United States and Canada, Australia and New Zealand, and few small Asian countries. Middle- income countries are the 72 countries with GDP per capita smaller than US$9000 and greater than US$735— these include countries such as China, Egypt, Brazil, Turkey, and so forth. Low- income countries are the 64 countries with GDP per capita smaller than US$735— these include countries such as India, Nigeria, Pakistan, and others.
One can take the total GDP of each category, divide by the total population of countries that belong to a certain category, and calculate the average GDP per capita for that particular group of countries. Table 9.1 shows the results together with the corresponding value for the entire planet. All of this is a snapshot of the present situation (2003). Here we attempt to make projections at least to the edge of “now.” Figure 9.4 shows some of the trends that need to be accounted for. The figure shows the changes in the GDP per capita since 1960 for India and the United States. India was chosen as an example because it is the most populous country in the low- income group while the United States was chosen because it is the most populous country in the high- income group. You will notice something peculiar in the graph not
Table 9.1.
Global income groups and average GDP per capita
| GDP/capita, US$ | |
| World | 5200 |
| High income | 26,900 |
| Middle income | 1900 |
| Low income | 450 |
Source: World Bank.5
Figure 9.4. Recent history of the GDP per capita of India and the United States in constant 1995 US$ Source: World Bank.5
yet mentioned in this book. The scale of the vertical axis is not linear: 100– 1000 is the same distance on the vertical axis as 1000– 10,000, which on a linear scale would take 10 times the space. Such a scale is called a logarithmic scale. It is used here because the Indian GDP per capita is more than a factor of 10 smaller than the US GDP per capita. If we showed both countries on the same linear graph, then the scale of the graph would have to be adjusted to that of the United States, and India will be extremely close to the horizontal axis. Plotting them on a logarithmic scale makes both economies visible and relatively easy to compare.
I will push this a bit further. Inspecting the two lines on this scale, we notice that the US line is approximately a straight line, while the India line changes to an increased slope around 1980. Straight lines on this scale means exponential growth with a constant growth rate (remember Chapter 2). One can calculate this rate from the graph. In this case I will skip the details. Th e rate for the United States is 1.5%. This rate is expressed in constant US dollars, which means that the inflation rate in the United States is discounted. We can do the same for India— the Indian economy grew until 1980 at an approximate rate of 0.9% in real US dollars, followed by an accelerated rate of approximately 2.5%. These calculations are performed with units that already discount population growth (because we calculate GDP per capita) and inflation in the United States. It is an old axiom of geometry that two unparallel, straight lines must meet somewhere. From the graph, we can calculate where the lines will meet. If the US economy continues to grow at the same rate that it grew since 1960, and the Indian economy continues to grow at the same rate that it grew since 1980, then the two economies will have the same GDP per capita in about 300 years. This time is way beyond “now,” but it is a trend that needs to be seriously considered. Presently the average American is richer than the average Indian by a factor of about 60. W hen the two extrapolated economies will converge in terms of an average individual standard of living 300 years from now, the GDP per capita of the two economies will reach the astronomical number of 3.5 million US$ (this is already aft er discounting inflation and population growth). Such is the power of an extrapolated exponential growth without imposed limits.
The third term in the IPAT equation is energy/GDP— how much energy a country needs to generate a unit of GDP. The lower the number, the more efficient the use of energy to enhance the standard of living. The common name for this is the energy intensity. In Chapter 2 we saw a graph of GDP versus energy use (Fig. 2.4). The data in that figure were a 1-year snapshot (2000) and show an approximate straight- line (linear) behavior with a great deal of scatter because for that particular year, the energy intensity was approximately constant, independent of how rich the country was or how much energy it used. I introduced the graph in Chapter 2 in order to emphasize why the use of fossil fuels became such an important policy issue. The perception is that use of low-cost fossil fuel energy directly leads to an increase in the standard of living. Th e opposite is also believed to be true— attempts to limit the use of fossil fuels as an energy source will directly affect the growth of economic activity. This is the main reason that the US administration in 2004 refused to sign the Kyoto Protocol, which attempts to limit the growth in GHG emissions—an issue that will be discussed in some detail in Chapter 13.
Figure 9.5 shows another aspect of this issue: the recent history of changes in energy intensity for India and the United States. The figure starts in 1975 where Indian energy use was about 25% more effi cient (i.e., the energy intensity was about 25% smaller) than that of the United States, and the figure ends in 2000 where the energy intensity of both countries converges. The converging point marks about a 25% decrease in the energy intensity of India and about a 38% decrease for the United States (i.e., an increased energy effi ciency for both countries). The reason for this increased efficiency is now a topic for a major research eff ort. Unquestionably, the reasons involve changes in the importance (or, to use a term that economists use, “weight”) of the various components used to calculate the GDP— changes away from energy- dependent heavy industry and agriculture and toward service and information industries, which are less dependent on energy. In Chapter 12 we will return to this issue to explore the sensitivity of our changes in energy use to the price we pay for energy.
Figure 9.5. Recent history of energy intensity of India and the United States Source: World Bank.5
In this chapter we have examined three concepts from the right- hand side of the IPAT equation. We found that the global growth in population is decreasing, with a predicted saturation of the global population at around the end of the century. Inspection of the data shows that the global standard of living as measured by the GDP per capita is increasing exponentially, with recent growth of the most populated developing countries outpacing the economic growth of developed countries. We found that the recent trends in energy intensity are down, which means we can produce more economic activity with a smaller input of energy globally.
The role of the last two terms of the equation and the probable limits to the trends we are discussing here will be explored in the next two chapters. At this stage I will develop a scenario of a fossil fuel– based economy in which the mix of fuels (coal, natural gas, and petroleum) does not vary too much across the world. In that case the last two terms will remain constant, independent of energy use, GDP, and energy intensity, and our attention will shift to the left - hand side of the IPAT equation.
For all the reasons that we have examined, our collective aspirations are to minimize emission of CO2 per capita and maximize the generation of GDP per capita. However, in order to do that we must first examine if the IPAT equation is just an attempt to separate the various reasons for generating GHGs so that the units on the right side of the equation agree with the units on the left side (dimensional analysis) or if the functional relation is actually supported by the available data.
Figure 9.6 shows the recent history of CO2 emissions for India and the United States. Th e CO2 emissions are divided by the population, thus moving the population term from the right- hand side to the left- hand side of the IPAT equation and, in so doing, normalizing to the increase in population of the two countries. The figure is also presented in linear scale, although the closeness of the Indian data to the horizontal axis makes it very tempting to resort to a logarithmic scale. In the figure we see a sharp increase in the US emission from 1960 to 1970 that, for reasons that will be explored in the next few chapters, levels off aft er 1970. The Indian emission level (per capita) is steadily increasing, although it is still lower than the US emission by more than a factor of 10.
Figure 9.6. Recent history of CO2 emission per capita of India and the United States Source: World Bank.5
Figure 9.7 shows (this time in logarithmic scale) the recent history of the product of the GDP per capita and the energy intensity of India and the United States. Under the assumptions that the last two terms of the IPAT equation are constant, if the IPAT equation can be supported by the data, then the product of the GDP per capita and the energy intensity should reflect the same trend as the CO2 emissions. Figure 9.7 shows that indeed this is the case.
If we decide to keep using fossil fuels as our main energy source, the feasibility of which and the possible alternatives to will be examined in the next two chapters, then the only way to minimize CO2 emissions and at the same time increase our standard of living (i.e., GDP per capita) is to decrease the energy intensity faster than the increase in the GDP per capita. Determining if there are upper limits to this process is an interesting problem— one in which I am trying to interest my economist friends with limited success.
As was discussed in the previous chapter (Fig. 8.1), the predictions of the anthropogenic contributions to climate change are based on the predictions of our socioeconomic development. The IPAT equation (equation 9.1) captures these developments as population, GDP, and terms that relate to energy use. Prediction of these developments is a difficult task I will revisit in almost every remaining chapter of the book. The IPCC knows outcome predictions cannot possibly be better than the input that goes into the computer models that calculates the outcome. This is shown clearly in Figure 8.1. The IPCC is addressing these issues by avoiding predictions. Instead the IPCC relies on a set of scenarios. Every scenario is in essence a possible story of what will happen with an associated outcome that produces exact numbers as to the environmental stresses that result. These numbers serve as the input to the computer models that calculate the climatic consequences of these scenarios. The scenario family includes 40 scenarios known as Special Report on Emission Scenarios (SRES).6 The IPCC emphasizes that it is not pretending to know what the future will bring and that the probability of any given scenario to materialize is the same as any other. Recently, I have tried to address the issue of whether, even in principle, we have a chance to avert a global disaster.7 In order to accomplish this I took two of the scenarios from the SRES compilation— one that represents a business- as- usual scenario and the other a friendlier (to the environment) scenario and superimposed these scenarios on recent changes in socioeconomic activities. The results in terms of projected growth of population, GDP, and emission are shown in Figures 9.8– 9.10. The results in terms of energy use will be shown in Chapter 11, where I discuss alternative energy sources.
Figure 9.7. Recent history of the product of GDP per capita and energy intensity of India and the United States Source: World Bank.5
Figure 9.8. Real and projected changes in global CO2 emissions in billion tons of carbon Source: Tomkiewicz (2010).7
Figure 9.9. Real and projected changes in global population Source: Tomkiewicz (2010).7
Figure 9.10. Real and projected changes in GDP per capita Source: Tomkiewicz (2010).7
The eye- catching aspect of this analysis is that although Figure 9.8 clearly shows that in terms of CO2 emission B1 is the environmentally friendly scenario and A2 is the business- as- usual scenario, Figure 9.10 shows that, on average, we are much bett er off with the environmentally friendly scenario. The reason is that the A2 scenarios predict a much faster population growth. However, in the beginning of this chapter it was shown that the median population growth scenario based on UN data, which has the best track record of accuracy, is very close to the B1 scenario, and we have discussed some of the driving forces predicted to continue to mitigate the population growth. One additional aspect of these figures is worth mentioning here—Figure 9.8 shows that the predictions for CO2 emissions until 2050 show very litt le difference; it seems we will get there no matter what we will do. Th e difference starts to materialize after 2050. Th is is my fork. It starts even earlier than my earlier definition of now. In Chapter 11 we will investigate the split in the two scenarios in terms of our energy use.
