The second session of the workshop was chaired by steering committee member W.G. Ernst (Stanford University). It included presentations by John Bongaarts (The Population Council), Andrew Jorgensen (University of Utah), and Peter Marcotullio (Hunter College). Their presentations are summarized here, followed by the discussion that ensued.
Vice President and Distinguished Scholar, Population Council
John Bongaarts noted the extraordinary demographic changes in recent history. The most obvious is the rise in population numbers (an increase of 5 billion since 1950), but he identified other important trends. There has been a revolution in the area of reproduction and fertility; the global fertility rate has been cut in half since 1950. People are also living longer and healthier lives; life expectancy is rising rapidly in almost all countries. Migration flows are expanding as well, as people move away from rural areas and venture across country borders in search of a better life. Also, populations in general are aging. Bongaarts focused on population trends in his presentation.
According to Bongaarts, world population was 1.6 billion in 1900, 2.5 billion in 1950, and has exploded (primarily due to a decline in the
FIGURE 3-1 Population estimates and projections from 1900 to 2100.
SOURCE: Bongaarts presentation, slide 3.
death rate) to 7.2 billion today. Current projections show that population is expected to reach just below 11 billion around 2100. (See Figure 3-1 for a plot of population by continent.)
Asia, which has tripled in population size since 1950, dominates this growth, but Africa is currently the most rapidly growing region and will be the largest contributor to population growth in this century. Europe’s population is expected to decline overall, with declining European nations outpacing growing ones. Bongaarts pointed out that the United Nations produces high-, medium-, and low-variant projections by changing the assumptions about future fertility rates. The population estimates may change in the different projections, but the overall percentages by region remain more or less the same. Figure 3-1 represents the medium values from the United Nations, and there is an uncertainty of 2 to 3 billion people associated with those total numbers.
Looking at the populations of each country, Bongaarts pointed to high growth in Africa and parts of the Middle East. In contrast, in North American and European countries, population growth is slow or declining. The main driver is variation in fertility. Over time, nations tend to move from high fertility and high growth to low fertility and low growth.
FIGURE 3-2 Scatter plot of nations’ growth rate as a function of fertility rate.
SOURCE: Bongaarts presentation, slide 3.
Bongaarts showed a scatter plot of growth and fertility levels (see Figure 3-2) and divided nations into one of three groups:
- Group 1: Characterized by rapid population growth and a young age structure. In sub-Saharan Africa, almost 70 percent of the population is under age 25. (In the developed world, by contrast, 30 percent of the population is under age 25.) Group 1 nations tend to experience depletion of natural resources; increased pollution; high unemployment; high maternal and infant mortality; lagging government investments in education, health services, and infrastructure; and rising political unrest and crime due in part to the unemployment of young people.
- Group 2: Characterized by lower fertility rates, lower population growth rates, and fewer young people. They are reaping a temporary “demographic dividend”—an economic boost from a rising proportion of workers. Bongaarts explained Group 2 nations tend to experience increased consumption, leading to increased pollution; rapid economic growth; declining maternal and infant health problems; government investment in education, health, and infrastructure; and rising inequality among their populations.
- Group 3: Characterized by low fertility rates, near-zero population growth, and a rapidly aging population. Bongaarts said Group 3 nations tend to experience high greenhouse gas emissions, the result of high levels of consumption; slower economic growth; slower gross domestic product (GDP) growth; lower growth in worker productivity; unsustainable pension costs; rising deficits; and rising voting power of the elderly.
These groups show distinctly different problems with different solutions; for example, Group 1 has problems due to high fertility, while in Group 3, problems are associated with aging. In high-fertility nations, Bongaarts indicated that the need is to invest in family planning programs and in human capital (particularly the education of girls). In aging nations, Bongaarts said the solution likely resides in a combination of modifying the pension system; encouraging a larger labor force, including working later in life; encouraging childbearing; and encouraging immigration.
Andrew Jorgenson, Professor and Director of Graduate Studies, Sociology
Department, University of Utah (presenter), and
Juliet Schor, Professor, Sociology Department, Boston College (co-author)
Andrew Jorgenson began by asking two questions: (1) What are the planetary boundaries and thresholds that would, if crossed, lead to unacceptable environmental change, and (2) What are the human drivers of consumption and environmental change that are pushing the population to and over the planetary boundaries? He said he would focus primarily on population and affluence.
Jorgenson defined drivers as the demographic, economic, and other social factors that put direct and indirect pressures on the environment. Drivers can be nested and aggregated. Jorgenson then looked at resource consumption and environmental change. To measure resource consumption, one can focus on direct consumption measures (such as food or other natural resources) or use a measure of consumption-based environmental demand (such as ecological footprint). Jorgenson said he likes the ecological footprint measure because it is measured in area units, not economic terms. Three possible metrics, each of which reveal something different, can measure carbon emissions: total carbon emissions, emissions per capita, and emissions per GDP. Depending on the stakeholder, the focus is on one measure over another. Jorgenson pointed out that scale-level
outcomes are likely the most critical to issues of sustainability. He showed time series figures to assess emissions in different points in time.1 In the early 1960s, the majority of nations had a per capita ecological footprint below the globally sustainable threshold. By 2003, more than half of them were above the sustainability threshold, although the distribution had broadened. In other words, in 2003, some nations were still underconsuming, despite the overall increased trend toward overconsumption.
Jorgenson then discussed affluence, noting that the traditional IPAT literature does not cover much discussion of heterogeneity in the effects of affluence or population size. He showed elasticity coefficients that result from statistical modeling of carbon dioxide emissions from the burning of fossil fuels, looking at the population effects on emissions. He used a two-way fixed effect model, which can be useful for hypothesis testing. In Asia, the elasticity coefficient2 is decreasing over time, dropping from 1.49 to 1.28 between 1960 and 2005. In Africa, the elasticity coefficient has also decreased long term, moving from 1.0 to 0.81. In the high-income countries, the elasticity coefficient is stable and large (around 2). As a result, any population growth in these countries could have a large effect on emissions. Jorgenson pointed out that the results are roughly equivalent when considering the ecological footprint instead of carbon emissions; the correlation factor between the two is around 0.9, but with both regional and temporal variation. He also considered the effect of GDP per capita on emissions by splitting countries into two categories (22 countries were considered developed, 64 less developed). He found very strong temporal stability in the elasticity coefficient for the less developed countries, but a slight decoupling between GDP and emissions for the developed countries beginning in around 1985. He speculated that this could be the result of technology improvements, as well as the influence of expanded global production and trade networks. However, he posited that it is difficult to measure and determine how such mechanisms might shape the relationship between emissions and affluence.
Jorgenson suggested that it would be helpful to have sustainability research that examines how to reduce the ecological or carbon intensity of human well-being in an integrated way. Ecological intensity of well-being is the level of environmental stress (for example, ecological footprint per capita) per unit of human well-being (for example, average life expectancy). The carbon intensity of well-being is the level of anthropogenic
1For Jorgenson’s presentation, see http://sites.nationalacademies.org/DBASSE/BECS/CurrentProjects/DBASSE_072678 [March 2014].
2The elasticity coefficient is the estimated percentage change in the dependent variable associated with a 1 percent increase in the independent variable, controlling for all other factors in the model. In this case, the independent variable is population and the dependent variable is carbon emissions.
carbon emissions per unit of human well-being. These two measures may be useful to provide a more integrative approach to the study of sustainability. Jorgenson showed data from a small sample of 14 high-income and 31 developing nations. Developed nations showed a sharp increase in the estimated effect of GDP on ecological intensity of well-being between the 1960s and 1980s, with the elasticity coefficient crossing from negative to positive around 1970. (He pointed out that, from a sustainability perspective, a negative elasticity coefficient when measuring the effect of GDP on the ecological intensity of well-being is desired.) The lower-income nations also saw an increase in the estimated effect of GDP on ecological intensity of well-being from the 1960s to the 1980s, but the overall value still remained near or slightly below zero. He stressed examining regional effects, but said the overall increasing trend between GDP and ecological intensity of well-being was a troubling one.
Jorgenson recognized the growing environmental concern among individuals in nations throughout the world and asked how to convert that concern into behaviors and actions, both individual and collective, to help address broader sustainability issues. He said that integrative approaches to sustainability studies are part of a critical path forward.
Peter Marcotullio, Associate Professor, Department of Geography, Hunter College of the City University of New York (presenter), and Karen C. Seto, Professor of Geography and Urbanization, Yale School of Forestry and Environmental Studies, Yale University (coauthor)
Peter Marcotullio began his presentation by making five statements: (1) Urbanization is a “mega-trend” for the 21st century; (2) there are challenges associated with environmental sustainability and urbanization; (3) there are many opportunities to gain efficiencies; (4) despite these opportunities, the scale of the issues is overwhelming; and (5) humans have the potential to create biophilic, resilient, livable cities in the future. Marcotullio noted global urban population is increasing and will continue to do so throughout the coming century. By 2050, rural populations will be stable or decreasing around the world. He pointed out that there are many estimates of current urban land use. One estimate finds that in 2000, 73 million hectares of land was urban; by 2030, those numbers may more than triple, to an estimated 225 million hectares (Seto et al., 2012). The change, both in land use and population, has a direct
impact on the environment, including lost habitat, lost agricultural land, and alteration of natural resources. It also has an impact on consumption of resources (wood, iron, etc.) to build cities. Indirect impacts from this growth are related to increased consumption and associated waste and emissions. Marcotullio said analysts often conflate increased consumption with urbanization itself, but he suggested considering these two issues separately. Energy consumption, he noted, varies as urbanization patterns evolve. Studies suggest a period of high growth and associated high-energy consumption early in the development process, followed by a period of slowing growth and associated decrease in energy use growth. He reported this pattern happened early in the United States and Japan. China is still in the initial growth period, although the trend in China may be tempered slightly; China’s energy intensity is already dropping, whereas during the same level of economic development in the United States, energy use was still increasing.
For the years 2005–2050, Marcotullio predicted enormous increases in consumption of certain materials (particularly metals), with a direct impact on populations as nations look outside their own borders to obtain needed materials. He projected rapid urbanization without the accompanying necessary basic infrastructure development (i.e., water and sanitation) in some parts of the rapidly developing world, noting 20 percent of the global urban population does not currently have access to an adequate water supply. The developed world will likely experience growing infrastructure deficits as older systems deteriorate. For example, the American Society of Civil Engineers released a report card about America’s infrastructure in 2013 (American Society of Civil Engineers, 2013) that examined such infrastructure categories as water, sanitation, dams, bridges, roads, and schools. The overall assessment was very negative with a “D+” grade.
Marcotullio stressed a number of opportunities along with the challenges. These opportunities include the following:
- “Agglomeration economies” in production and consumption.
- Lower energy emissions per capita through consolidation. Marcotullio explained that this has been true in the United States, but not necessarily yet true around the world. In some cases, cities provide additional access to energy systems, which enhance consumption as compared to rural areas, which overwhelms the efficiency effects of density (Marcotullio et al., 2013).
- Scale economies in production and consumption.
- Increases in labor productivity.
- Consumption amenities.
- Increase in technology and innovation. Marcotullio showed that there is a disproportional increase in patents and other measures of innovation with increasing city size;
- Reduced transportation emissions. Increased residential density lowers transportation emissions anywhere from 5 to 25 percent, and the benefit tends to increase over time. Marcotullio stated, for example, that bicycle sharing can be a successful model; in New York, for example, 100,000 customers signed up in the first 10 days of the program.
Marcotullio cautioned that, at the same time, efficiency gains may be overshadowed by the scale of urban expansion, as seen in the “rebound effect” in the United States after the 1970s oil shortages. During the late 1970s and early 1980s, transportation figures demonstrated that motor vehicle drivers dramatically reduced fuel consumption through higher efficiency cars and reduced driving. However, these gains were lost over time, and by around the turn of the century, motor vehicle fuel consumption per capita had reached new highs. Any gains from greater fuel economy were lost as more people drove more miles. More directly, Güneralp and Seto (2012) pointed out that in the Pearl River Delta, China, efficiencies gained in energy consumption from dense settlement are overwhelmed by the absolute scale of urbanization. Marcotullio ended on a note of both optimism and urgency, pointing to the opportunity to make cities more environmentally benign (i.e., biophilic, resilient, low carbon, and livable) because 65 percent of the global urban environments (defined by urban land cover) predicted to be developed by 2030 have yet to be built.
The discussion session opened with the topic of population targets. A participant pointed out that demography is not destiny, and there is an uncertainty range associated with the various demographic projections. Another asked if there is a desired, or target, world population for 2050, and if there is an optimal level of fertility to help counteract the effect of aging populations. Bongaarts responded that he would be very concerned to see very low fertility rates (1.5 or below). He suggested that something near 1.8 is optimal for societal stability, but government intervention may be necessary to attain such a value. Bongaarts stressed that aging societies remain a huge demographic problem. While working later and postponing retirement is part of the solution, he said these actions alone cannot resolve the large imbalances in social security and health that would occur with an aging population. He said aging populations are among
the biggest policy and economic challenges facing the United States and Europe in the next two to three decades. Marcotullio discussed the dividend-burden movement; infrastructure is often built when population growth is taking place, and builders do not think about the future needs of an aging population. For instance, Tokyo is built with many stair steps needed to cross a street, something the elderly can have difficulty using. Jorgenson also stated an increased-age population structure has an impact on energy consumption.
Another participant asked about the differences across nations in birth rate reduction, such as the drastic decline in Mexico. Bongaarts said two things are needed to reduce birth rates: an educated population and access to contraception. Iran, for example, made a large investment in schooling, but the fertility rate remained around 6 births per woman. In 1989, the Iranian government examined projections of infrastructure requirements needed to support its burgeoning population and realized that a crisis was looming. The Iranian government instituted a family planning program that provided many options and was accessible to rural populations. Fertility rates in Iran dropped from 6 to 2 within a decade. Bongaarts said Mexico similarly developed a strong national program in health and family planning. He pointed out that the strength of the economy does not influence fertility rates in a significant way. Nigeria, for example, has a healthy economy but still maintains very high fertility rates. A participant asked about the different impacts of the role of women in the workforce. Bongaarts replied that the child-bearing and child-rearing years can take up the entire life of a typical African woman: She can have children from ages 18–40 and would be 60 before the youngest children left her home. A German woman, by contrast, more likely has two children very close together in age. She would be absent from the workforce for only a few years.
A participant asked about the quality of migration models. Bongaarts responded that migration is modeled rather poorly. He said migration is a minor factor in most population estimates, but it is very important to small, focused regions, particularly some areas in South America. United Nations models assume that migration will go to zero in a few decades, but he said he does not find that to be a realistic assumption and predicted the topic will increase in importance. Recently, there have been examples of rapid migration in Europe and, to a certain extent, in North America, and he said populations that immigrate tend not to integrate quickly into their new society.
Marcotullio postulated that identifying a person’s location is a difficult task, and that spatial distribution of population, including migration, is a new demographic area that needs additional work. Jorgenson pointed out that migration may be important in understanding envi-
ronmental change; a recent task force on climate change is showing that migration is a key knowledge gap. However, Bongaarts said he does not consider migration to be a factor in which large or surprising outcomes are likely. He said that instead there might be more room for variation in mortality rates and health improvements. On the other hand, he noted, a large airborne virus could have surprising outcomes on population. The population projections in the 1990s did not include the significant drops in fertility rates since observed. Overall, however, Bongaarts expressed confidence in current median projections of population.
Several participants asked about the definition of “urban” and whether to revisit that definition. One participant pointed to a division of the term into two definitions: an entity-based definition (population, density, land use, etc.) and a quality-based definition (the “urbanness” of the life experience). Marcotullio said he used an entity-based definition in his presentation. However, population densities are decreasing, and the question arises at what density an area is no longer considered “urban.” Turner also asked about differentiation and urbanization by level of development. Marcotullio responded that the economics of agglomeration hold true, but the impact on the environment will differ by the level of development. In developed nations, such as the United States, emissions per capita are lower in cities because of agglomeration. Per capita emissions are higher in cities in other nations, particularly developing nations where there is a disparity in access to resources between urban and rural populations. As resources diffuse across populations, per capita emissions in cities are likely to be lower than per capita emissions overall.
A participant turned the discussion to economics, asking about the demographic dividend and its relationship to consumption and other demographic variables. Bongaarts said, to an economist, the demographic dividend was only positive because it results in a decline of the poor populations and motivates governments to help decrease fertility rates. The downside is that the effect lasts for only a few decades, and then it shifts to the demographic burden of an aging population. A participant asked whether the demographic dividend is actually an education dividend. Bongaarts said that a boost in education tends to be the result of a population decline; it is easier to educate fewer students.
Another participant asked about the robustness of the ecological footprint as an indicator, suggesting it might be useful to correlate to carbon dioxide emissions, but not for other variables. Jorgenson agreed it is right to be skeptical of the ecological footprint as an indicator and said he employs it increasingly less often. He noted he likes to use it for group discussions and in classroom settings, as the ecological footprint concept brings the issues to the table in an understandable way.
The topics of livestock and food behaviors were brought up (also
discussed in the context of resource distribution in a later session). A participant posited that demographers tend to be concerned with drivers of emissions that include methane, so livestock is considered from that rather narrow perspective. However, another participant referred to a 2006 report by the Food and Agriculture Organization (FAO) of the United Nations, which stated that the total environmental impact of livestock is larger than that of the transportation industry (Steinfeld et al., 2006). Providing protein sources to a large population may be an important topic, the participant said.
This page intentionally left blank.