Intermediary Cities and Climate Change

An Opportunity for Sustainable Development

Report

16 November 2022

2. The linkages between climate change and intermediary cities

Abstract

The objective of this chapter is to reduce the knowledge gap on the linkages between climate change and intermediary cities. Although knowledge about climate change has seen constant increase over the last two decades, little is known about how it is affecting small and medium-sized agglomerations in developing countries. Such cities are playing a key role in urbanisation in regions like Southeast Asia and Sub-Saharan Africa. However, due to socio-economic, institutional, and geographical issues, these intermediary cities are disproportionally exposed to climate change.

Climate change and fast urbanisation are arguably among the biggest challenges facing the world today. Climate change is reshaping our livelihoods and increasingly exposing large shares of the population to unprecedented risks. The Sixth Assessment of the Intergovernmental Panel on Climate Change (IPCC) warns that global warming is expected to reach or exceed 1.5°C on average over the next 20 years, and possibly by the early 2030s, ten years ahead of previous projections (IPCC, 2021[1]).1 As such, many ecosystems will shortly reach tipping points that lead to irreversible damage, while socio-economic costs will continue to increase. Along with the looming climate crisis and ecological breakdown that we are witnessing today, human settlements in developing countries are also rapidly transforming. The urban population in many countries in Sub-Saharan Africa and Asia will double in the next three decades (UNDESA, 2018[2]), while in Latin America several cities will struggle to cope with high levels of congestion and increasing inequality. While urbanisation can present large benefits for socio-economic gains and rural-urban transformation (i.e. economic transformation in rural areas triggered by urbanisation, such as poverty reduction, changes in food consumption and production, etc.), it can also present systemic challenges that can further expose urban dwellers to climate shocks and promote carbon lock-in. As such, managing the adverse effects of climate change on urban areas, while promoting low-carbon standards, is an urgent issue for local, national and international agendas.

Although intermediary cities are key for addressing climate change, they are often overlooked in both the international development agenda and the climate debate. This report considers “intermediary cities” by the key intermediation role they play in connecting metropolitan centres with rural areas, and for practicality they can be defined to be urban centres with a population of 50 000 to 1 million. In developing countries, these are dynamic agglomerations that connect rural and urban areas, as well as cities of different sizes within urban systems. They are usually small and medium-sized urban areas, i.e., not capitals or large cities that are strongly linked to rural economies. In many cases, intermediary cities have the potential to foster well-being, promote job creation and contribute to poverty reduction. They can also act as powerful laboratories for piloting new solutions and fostering social and economic innovation. However, many of these agglomerations are also characterised by limited capacity, poor governance frameworks and weak urban planning. These issues, paired with rapid urbanisation, hinder the capacity of authorities to provide public services and an environment safe from increasingly frequent extreme climatic events. Moreover, as these cities grow and become wealthier, they tend to move towards carbon-intensive energy sources, further contributing to greenhouse gas (GHG) emissions and climate change. For these reasons it is urgent to include the effects of climate change on intermediary cities in the development agenda.

There is a significant knowledge gap surrounding the nexus between intermediary cities and climate change. In studies of climate change and cities, large or capital cities have received more attention. Overall, for each publication about intermediary cities or small and medium-sized cities, there were close to 142 publications studying big cities (including scientific articles, peer-reviewed articles and reports from development institutions). In the case of climate change the gap narrows, but there are still ten publications about climate change and big cities per each publication about climate change and intermediary cities.2

To address this gap, this chapter uses novel datasets in order to better understand the relationship between climate change and cities of different sizes. In particular, it uses the Urban Centres Dataset developed by the European Commission and the OECD. As such, the chapter analyses the urban dynamics characterising intermediary cities and their multidimensional vulnerabilities to climate change, as well as their contribution to greenhouse gases.

The chapter highlights that there is no simple narrative linking intermediary cities and climate change. It shows that the effects of climate change, and the contribution to greenhouse gases, are not homogenous across the urban system. This is not linked to city size itself. Instead, it depends on geographic factors, such as distance to low-lying coastal zones, location in relation to the equator, etc.; the capacity of authorities to limit climate vulnerabilities and emissions; and governance mechanisms that ensure a sustainable path. Many of these features are not present – yet – in intermediary cities. By defragmenting knowledge and presenting new evidence, the chapter aims to promote future discussion of the management of urban systems in developing countries. It also aims to convey an optimistic message: the small(er) size of many intermediary cities could actually be a lever for achieving a sustainable transformation that avoids locking these agglomerations onto an unsustainable path, as has been the case for many larger agglomerations in developing and developed countries.

Why do intermediary cities in developing countries matter for climate change?

Climate action needs to take place now, and urban areas can play a critical role in both adaptation and mitigation efforts. Increasing evidence, including the latest assessment of the IPCC, reaffirms that global anthropogenic greenhouse gas emissions are not decreasing at the expected pace. Thus, the world is reaching a tipping point that could lead to irreversible damage to ecosystems and our societies.

As urban areas continue to grow, cities in developing countries are increasingly exposed to climate change. Urban areas in developing countries not only host a high share of the population but are also centres of economic activity, as well as nodes in complex networks of infrastructure including transportation, communication and information. Cities are also critical hubs for the provision of basic public services (health, education, etc.) that benefit not only urban dwellers but also the population of neighbouring rural areas. Events associated with climate change – such as floods, changes in precipitation, rising sea levels, etc. – threaten the intricate functioning of urban areas.

Until recently, limited attention has been paid to the role of intermediary cities in the international debate concerning climate change. There are different reasons for this, including a strong bias towards capital cities, limited available data on smaller agglomerations and an overall development narrative that has placed significant weight on the potential of large urban areas for national economic growth. However, national governments and international development partners are increasingly recognising the potential of intermediary cities for improving local well-being in developing countries. Due to their intermediation role, these cities act as hubs for the provision of goods and services, facilitate rural-urban migration and provide a conducive environment for income diversification and poverty alleviation.

Intermediary cities can also play an important role in building resilience to and mitigating the effects of climate change:

Intermediary cities have a major role in the urbanisation of developing countries. Intermediary cities will continue to host more than half the urban population of developing countries in the coming decades (UNDESA, 2018[3]). Moreover, intermediary cities in developing countries are continuing to expand, and in some cases very quickly. This has made these cities prone to losing density compared to metropolitan areas and capital cities. Urban processes underpinned by low density and fast expansion risk exposing vulnerable urban dwellers to the effects of climate change, while leading urban areas to emit higher levels of GHGs. This results from longer commuting times, increased use of motor vehicles, higher congestion, and reduced provision of services.
The intermediation role of these cities links their economies and societies to rural areas. By definition, intermediary cities have closer socio-economic and geographic linkages with rural areas. Intermediary cities act as nodes that connect people and socio-economic activities across rural and urban territories. As such, changes in climatic patterns that reduce agricultural productivity, increase the pressure on natural resources and limit rural livelihoods indirectly affect intermediary cities. These effects spread through different channels, two of the most important being food systems and internal migration. In other words, intermediary cities face the compounded effects of climate change: a direct effect resulting from more frequent extreme climatic events, and an indirect effect resulting from disruptions in food systems and historic migration patterns. Similarly, climate effects on intermediary cities largely affect rural areas and the well-being of their populations. It is therefore important to understand their intricate relations and the effects of climate change on their mutual linkages.

Due to the complexity of the indirect effects of climate change on intermediary cities, this topic is the main focus of Chapter 3, while being briefly explained in this section. Box 2.1 outlines the definition of intermediary cities for this report and highlights the three constraints they face: knowledge gaps, policy gaps and financing gaps.

Box 2.1. What constitutes an intermediary city?

Intermediary cities are urban agglomerations that play an intermediation role that allows them to connect metropolitan and rural areas as well as different groups of cities within urban systems. There is, however, no universal definition for intermediary cities. Although intermediary cities are characterised by their intermediation role, there is no consensus on how to identify this role based on their main features. Moreover, there is limited data on the economic activities, employment share and economic specialisation occurring among these cities in developing countries.

For this reason, in this report, intermediary cities are considered to be those urban centres with a population ranging between 50 000 and 1 million inhabitants.

The main benefit of this approach is its simplicity. This is a clear rule that can be easily applied to existing information. Moreover, across developing countries, it does a good job separating small and medium-sized cities, which usually play an intermediation role, from large and capital cities.

There are, however, a number of caveats, the most important being that this approach does not rely on the function(s) of a given city within the urban system. In addition, it does not take into account the fact that highly populated countries tend have highly populated cities, meaning that in some countries this threshold may be too low. This is probably the case for countries like China and India.

A methodology for identifying intermediary cities in developing countries is envisaged as one potential output of the Development Centre during the next programme of work.

Intermediary cities play a critical role in the urbanisation of less developed regions

Urbanisation is a global phenomenon. In 2020, 55% of the total global population resided in urban areas (UNDESA, 2018[2]). However, urbanisation rates differ across regions. For instance, North America is the most urbanised region in the developed world, with 82% of its population residing in urban areas, followed by Europe (74%) and Oceania (68%). Latin America and the Caribbean (LAC) is the most urbanised region in the developing world, with 81% of its population residing in urban areas, followed by Asia and Africa, with 50% and 43% respectively (UNDESA, 2018[2]).

Urban populations are expected to grow very rapidly in emerging regions. Countries in Africa and Asia are urbanising most quickly. For instance, between 2018 and 2050, countries like China, India and Nigeria are expected to grow, respectively, by 255 million, 416 million and 189 million urban dwellers; these three countries together will account for almost 35% of total urban population growth around the world (UNDESA, 2018[2]). Figure 2.1 shows the urbanisation dynamics for developing countries in Asia, Africa, LAC and Oceania. The figure shows that, although China and India are – and will remain – the countries hosting the largest urban populations in the next three decades, they are not the ones expected to grow most quickly. Indeed, the fastest growing urban populations are based in Africa. The urban population of countries like Niger, Burundi, Uganda and Malawi will grow on average at annual rates above 4.5%. This implies that by 2050, these countries will be significantly more urbanised, and some of them may double their respective shares of urban population compared to their levels in 2018.

Figure 2.1. Urbanisation rates across developing countries

Intermediary cities account for a large share of the urban population in developing countries

Most urban dwellers in less developed regions will reside in small and medium-sized cities by 2035. Big cities are constantly in the spotlight due to their strong potential for generating economic growth, their function as hubs for knowledge and culture, and because they host large shares of national populations. Alongside large cities, urban areas with fewer than 1 million inhabitants are expected to continue hosting an important share of the world’s urban population in the next decades. Figure 2.2 shows the distribution of urban population across cities of different sizes in less developed regions. It shows that, in 2020, cities with fewer than 1 million inhabitants accounted for 58% of the urban population in less developed regions. Although this portion is expected to decrease in the next couple of decades, by 2035 these cities will still account for close to 53% of urban population in developing countries. The fact that small and medium-sized cities will host a significant share of the urban population in developing countries makes them key actors for climate adaptation and mitigation. Growth dynamics also imply that many large cities of the future are currently intermediary cities. This calls for urgent action now in order to ensure a sustainable growth pattern.

Figure 2.2. Share of urban population by city size, 1990-2035

How have intermediary cities grown in the last decades?3 Figure 2.3 shows average yearly growth rates of both population and built-up land across cities of different sizes between 2000 and 2015. Two important things to note:

The population of intermediary cities has not increased fastest across the board. Cities with a population of less than a million have experienced the highest population growth rates in Western Asia, Northern Africa and Latin America and the Caribbean (left-hand side of Figure 2.3). In contrast, cities with more than 1 million inhabitants grew fastest in Sub-Saharan Africa, South Central Asia and Southeast Asia. However, this does not imply that attention should be limited to larger agglomerations in the latter regions, since intermediary cities have also grown very fast in some of them. In Sub-Saharan Africa, agglomerations with a population of 500 000 to 1 million experienced an average growth rate of 3% over the period. At this growth rate, the population of these cities will double in 25 years. For local authorities this implies a massive mobilisation of resources in other to provide services and infrastructure to twice as many citizens.
But intermediary cities are growing fastest in terms of the expansion of built-up areas. Indeed, except for Western Asia and LAC, cities with fewer than 500 000 inhabitants experienced higher growth rates of built-up land than bigger cities across all regions (right-hand side of Figure 2.3). This is particularly the case for cities with 50 000 to 100 000 inhabitants in Sub-Saharan Africa, where average annual growth during the period reached 1%.

Across the world, urban expansion has become a growing concern. The IPCC (2014[4]) estimates that during the first 30 years of the 21st century, urban land cover will grow faster than what we have witnessed so far, approximately twice as fast as urban population growth. Asia is expected to account for almost half of the growth in global land cover. China and India are expected to account for 55% of the total regional land-cover increase, with China alone accounting for 18% of the total global increase in land cover (IPCC, 2014[4]). Urban expansion is expected to be most rapid in Africa, with almost 600% more urban land cover in 2030 compared to 2000 levels (Seto et al., 2012[5]). As will be highlighted in later sections, uncontrolled urban expansion can have major negative implications in terms of both climate mitigation and climate adaptation, and it particularly poses large risks for rapidly growing small and medium-sized cities.

Figure 2.3. Average growth rate of urban population and built-up land by world region, 2000-15

While the previous figures highlight trends across major developing regions, they mask important differences. Figure 2.4 compares the growth rates of population and built-up areas of intermediary and non-intermediary cities in developing countries for the period 2000-15. To account for national differences, both growth rates are expressed as standard deviation from the national mean. In other words, observations close to zero – on either axis – represent cities with growth rates close to the national mean; in contrast, observations with values closer to 3 (or -3) represent cities with growth rates significantly higher (or lower) than the national average. The figure shows high levels of dispersion for growth rates of both population and built-up land. However, in the case of built-up areas, cities with more than 1 million inhabitants show a strong tendency to concentrate below their country average, while cities with fewer than 1 million tend to be more dispersed, surpassing 2 standard deviations. As such, cities with fewer than 1 million inhabitants tend to have high built-up growth rates that are at times paired with high population growth rates and at others with low population growth, in which case cities tend to lose population density.

Figure 2.4. Dispersion of population and built-up growth rates from 2000 to 2015

Intermediary cities lose density when built-up expansion outpaces population growth

Many small and medium-sizes cities have lost density since 1990 because the rapid expansion of built-up areas has outpaced population growth. Figure 2.5 shows the percentage of cities in developing countries that lost population density during the period 2000-15. This drop in density was most acute in intermediary cities. For instance, around 63% of cities with 50 000 to 100 000 inhabitants lost density over the period, compared to only 31% of cities with more than 1 million inhabitants.

Despite the high heterogeneity of growth patterns in intermediary cities, and once we control for cities and countries’ characteristics, small(er) size is associated with decreasing population density, according to results from econometric analysis for the 2000-15 period. Indeed, the odds of density decreasing in cities of 50 000 to 100 000 inhabitants are around 5.4 times higher than in cities with more than 1 million inhabitants (holding all variables constant); almost 3.5 times higher for cities with population of 100 000 to 500 000 inhabitants; and 2.3 higher for cities with population of 500 000 to 1 million inhabitants (Figure 2.6). In contrast, being a coastal city decreases the odds by 40-50%, holding all variables constant.

Figure 2.5. Percentage of cities where density decreased over 2000-15

Figure 2.6. Results from logistic regression analysing density decrease

How cities develop has a major effect on their climate resilience and GHG emissions

Declining population density is strongly associated with urban sprawl. Yet urban sprawl is a complex phenomenon: the term is used inconsistently, which can lead to confusion (Bae and Richardson, 2004[7]; Rubiera-Morollón and Garrido-Yserte, 2020[8]). For a working definition of urban sprawl, this section uses Bae and Richardson (2004[7]). The authors define urban sprawl as “a pattern of land-use in an urban area that exhibits low levels of some combination of eight distinct dimensions: density, continuity, concentration, clustering, centrality, nuclearity, mixed used and proximity”. They also link the concept to the notion of automobile dependency.

Sprawling and car-dependent cities can be referred to as high-demand systems with limited scope for a climate-sensitive development model. These systems are difficult to decarbonise and tend to perform poorly in terms of well-being outcomes like health, access to opportunity or safety. They can also lead to high GHG emissions and limit the potential of public policies and investment (e.g., in public transport) to improve such outcomes. Sprawling urban areas are often characterised by single-family dwellings -- the most common type of construction in suburban areas. Single-family dwellings are high-demand in terms of space, consumption of materials and heating, and are often less efficient than the multistorey buildings found in compact areas.

Urban sprawl, with associated declining population density, leads to higher GHG emissions as it tends to transform land use and shape commuting patterns. The expansion of built-up areas implies that there is less land covered by vegetation, and thus less absorption of CO₂ into the soil (UN-HABITAT, 2011[9]). Moreover, land clearing and deforestation for urban expansion lead to CO₂ emissions (Seto et al., 2012[5]). Another study, by Burgalassi and Luzzati (2015[10]), found that urban sprawl in Italy is associated with higher PM2.5 emissions from residential heating and CO₂ emissions from the transport sector. Furthermore, and despite limited evidence, there appears to be a correlation between the form of cities and energy consumption. Residential energy consumption tends to be higher in low-density urban areas, as suburban homes tend to be large and detached, requiring higher energy for heating than urban apartments (Litman, 2014[11]). Similarly, land consumption per capita is higher in low-density urban areas – and by reducing the amount of land available for agriculture, increased land development reduces overall agricultural production (Litman, 2014[11]).

Sprawling cities, along with scattered and single-use development, also mean that people need to travel long distances to meet almost every need (work, school, the grocery store, etc.). In this context, people tend to depend on cars (or other motorised private vehicles) to access places and opportunities, since active and shared modes of transport are no longer a convenient option for them (if an option at all) (OECD, 2021[12]). A sprawled urban system thus encourages, and many times imposes, car-centred mobility, which is a high-demand type of mobility in terms of space, fuel consumption, material demand, emissions and cost, both for individual users and governments. In other words, sprawling low-density cities lead to higher GHG emissions, higher commuting time and energy use, lower public transportation services (OECD, 2018[13]) and higher traffic congestion and air pollution (NCE, 2018[14]).

Sprawling cities with low density also impose higher economic and social costs. They lead to a dispersion of economic activities, which in turn leads to lower accessibility, especially for the poorest and most vulnerable urban dwellers, as space for walking and cycling as well as public transit is reduced (Litman, 2014[11]). Additionally, sprawl increases the cost of public service delivery for local authorities (NCE, 2018[14]), which in practice leads to lower service provision and higher dependency on private transport (OECD, 2018[13]). In fact, urban sprawl is conservatively estimated to account for a loss of 7% in the annual GDP of the United States (NCE, 2018[14]), reflecting a yearly cost of USD 400 billion in reduced public health and fitness and of USD 625 billion in costs borne by commuters (Litman, 2014[11]).4

In terms of well-being, evidence suggests a negative correlation between urban sprawl and environmental, economic and social outcomes, including health, housing affordability, access to opportunities, cost-effective public services, resilience and environmental sustainability (OECD, 2021[12]). Sprawling, car-dependent territories significantly constrain the scope of policies for improving well-being, with policy makers sometimes forced to make politically unpalatable trade-offs. For example, in sprawling urban areas where distances to places that people need to reach are long and alternatives to the car are not available or convenient, reducing emissions via policies that discourage car use (e.g., carbon prices) can result in lower quality of life, and thus be politically unattractive, if at all feasible. Box 2.2 describes the underlying factors that led cities in OECD countries to develop in an expansionary manner and the implications of this development on well-being.

Box 2.2. How policy choices fuelled urban sprawl in OECD countries

The drivers underlying urban sprawl have been widely analysed in the literature. They include land value (Pendall, 1999[15]), demographics (e.g. young couples with children seeking affordable housing) , land financialisation (Savini and Aalbers, 2016[16]), rising incomes that allow people to live in bigger homes in low-density areas (Carruthers and Ulfarsson, 2002[17]), decreasing commuting costs (OECD, 2021[12]; Brueckner, 2000[18]) and racial strife (Daniels, 1999). Many these drivers may be perceived as independent, individual preferences. For example, it is often perceived that people choose to live in bigger houses with gardens and to buy a car as soon as their income allows them to.

However, a systemic analysis of urban areas indicates that such choices are not solely the result of personal preferences but are instead largely determined by policy choices, and to a large extent by transport and urban planning policy choices, which are intrinsically linked. For instance, OECD (2021[12]) finds that decades of transport policies focused on mobility and privileging road construction have led to the dynamics of induced demand, urban sprawl and the erosion of sustainable transport modes. Induced demand is the phenomenon by which investments in road expansion to reduce congestion end up increasing congestion because the more roads there are, the more the use of cars becomes attractive, leading more people to choose to drive. The dynamic of urban sprawl in OECD (2021[12]) refers to the phenomenon by which people move further away from cities as road infrastructure allows them to reach places of interest (often concentrated in city centres) within a reasonable time, e.g. 30 minutes by car. The more roads expand, the more this is possible, illustrating how transport policies focused on road construction can incentivise choices leading to scattered and low-density development. In turn, induced demand and urban sprawl lead to the erosion of sustainable modes. As distances increase, active modes such as walking, cycling or micromobility are no longer an option. Moreover, as density decreases and single-use development expands, public transport is also less of an option, as it is difficult to get good service. These modes are, as a result, not competitive vis-à-vis the car.

Combined, these dynamics are at the source of car-dependent and expanded territories. They also determine the reach and effectiveness of public policies, and thus the capacity of authorities to improve people’s daily lives sustainably and cost effectively (e.g. the road infrastructure costs of car-dependent and scattered territories can be significantly higher than in more compact areas).

While the policy focus of the last decades has locked numerous countries into high-demand systems with poor well-being outcomes, the analysis above implies that urban development can take a different path and that climate policies can be aligned with other agendas. Understanding the dynamics underlying car dependency and expansionary development is fundamental to identifying which policies foster such vicious cycles and which could instead reverse them and lead to more virtuous cycles.

Developing countries have an enormous opportunity to avoid reproducing the patterns of urban development in Western countries. In this sense, the notion of leapfrogging, which is often limited to refer only to technology transfers from developed to developing countries, can be expanded and applied more holistically.

Low economic development increases the climate vulnerability of intermediary cities

What do we mean by climate vulnerability? Gasper et al. (2011[19]) highlight the nuances between climate vulnerabilities and climate risks. Climate vulnerabilities refer to the aspects or characteristics that render certain groups, individuals or cities vulnerable to climate-induced risks. These often include the socio-economic dynamics of the city, as well as the level of development and infrastructure. Climate risks refer to the likelihood of hazards affecting certain cities or areas due to their geographical location or to overall changes in climate (Gasper et al, 2011[19]).

The main underlying factors contributing to the vulnerability of urban dwellers in less developed regions are linked to the socio-economic dynamics of their cities. Population size and composition, geography, spatial development patterns, economic structure, inequality levels and the extent of informality are determinants of the resilience of urban dwellers to climate risks. In particular, what leads to losses in the economy, lives and livelihoods – and limits the ability of poor households and local communities to respond to and recover from disasters – are low income per capita, weak institutions and low adaptive capacities at household level and across different levels of government (Filho et al., 2019[20]). Most importantly, the availability and quality of early warning systems and effective post-disaster relief are essential for reducing the impact of climate hazards on the most vulnerable groups (Revi et al, 2014[21]). Reduction of climate risks largely depends on local governance capacity, i.e. the level of available human and material resources and technical know-how as well as the planning capability of local governments (see Chapter 4). This differs highly across countries. For instance, countries like Thailand and Bhutan have multilevel governance that makes them more resilient to climate risks than Myanmar, parts of Pakistan and Bangladesh, which are more vulnerable due to limited governance capacity (Busby et al., 2018[22]).

Climate change acts as a ‘’threat multiplier”. It increases cities’ socio-economic vulnerabilities, such as income inequality, resource depletion and poverty. Other environmental problems, such as air and water pollution, poor waste management and limited sanitation services, are exacerbated by climate change. This is particularly pertinent to cities in developing countries, where a large number of inhabitants live in poor-quality housing and informal settlements, heightening the vulnerability of the urban poor.

Climate change also presents health-related risks that are specific to the urban fabric. A rise in the global mean temperature of more than 1.5°C will intensify the effects of urban heat islands (UHI) as well as urban pollution levels (Mika et al., 2018[23]). It will also cause more frequent and intense heat waves in cities. A global mean temperature increase of 1.5°C will double the number of large cities experiencing heat stress by 2050, exposing approximately 350 million additional people to heat waves (Hoegh-Guldberg et al., 2018[24]; Ebi et al., 2018[25]). Inevitably, at 2°C warming there will be a stronger rise in heat-wave frequency and intensity (Mika et al., 2018[23]). By 2050, such a rise in global temperatures would leave more than 1.6 billion urban dwellers living in 970 cities – 45% of the global urban population – exposed to heat waves (UCCRN, 2018[26]).5

Infrastructure and service provision will be affected by climate change. As centres of economic activity, cities are highly dependent on complex networks of transportation, communication and information. Moreover, cities are hubs for the provision of social infrastructure such as health, education and administrative services to urban and surrounding rural dwellers. Public services and economic activities are interdependent and linked through various and complex networks, which makes them particularly vulnerable to climate-induced disruptions. Weather shocks can cause serious physical damage that disrupts both urban and rural economies (OECD, 2010[27]). For instance, floods can cause shortages in energy supply and damage to roads, transportation systems and water treatment centres. These disruptions affect not only the population but also key industries. In 2004, damage resulting from the flooding that hit Dhaka, the capital of Bangladesh, affected more than 681 garment factories and caused losses estimated at USD 9.1 billion. Similarly, floods in Dhaka in 1998 caused USD 30 million USD in losses for large firms and USD 36 million for small and medium-sized firms (Gasper et al, 2011[19]). Damage from weather shocks also affects rural areas surrounding cities since these climate events disrupt agricultural value chains and interrupt the provision basic services, which in turn reduces agricultural productivity and can lead to food insecurity and a decrease in rural income (Revi et al, 2014[21]).

Cities in Asia and Africa will be particularly affected by climate change

Cities in Asia and Africa are among the most vulnerable to climate change. According to the Maplecroft Climate Change Vulnerability Index, 95% of the total 234 cities identified in 2018 as facing extreme climate risks were located in Asia and Africa. In Africa, these cities included capitals such as Kampala (Uganda) and Addis Ababa (Ethiopia), as well as cities that serve as commercial hubs like Lagos (Nigeria) (Marplecroft, 2018[28]). Cities across Asia face different types of risks, depending on their geography and the capacity of national and local institutions. Cities located in low-elevation areas such as Myanmar, Bangladesh and northwest and southeast India are particularly exposed to cyclones and flooding, while cities in other parts of Asia – like northern Pakistan, Thailand, Sri Lanka and Cambodia – face water shortages and irregular precipitation (Busby et al., 2018[22]).

Similarly, climate risks differ in Africa according to the geographic location of cities. For example, precipitation is projected to increase in cities in eastern and western Africa, while cities located in coastal areas or along rivers will face increased flooding. This was the case for Maputo6 (Mozambique), where floods in 2019 caused 45 deaths (Cambaza et al., 2019[29]), or Bamenda (Cameroon) where 25% of the city’s 250 000 residents live in flood-prone areas (Tume et al., 2019[30]). Cities in western, eastern and northern Africa also face a high risk of more frequent and more intense heat waves. A study by Rohat et al. (2019[31]), assessing more than 150 large cities across 43 African countries, projects that by the end of the 21^st century (2090) the number of people facing dangerous heat conditions will increase 20 to 50 fold compared to 2019 levels (Rohat et al., 2019[31]). North African cities will face increasing desertification and a rise in average temperatures of 1.5°C to 3°C by 2060 (Rohat et al., 2019[31]). Rising sea levels will affect other cities, such as Mombasa (Kenya), which is projected to lose 17% of its land area due to a sea-level rise of 0.3 metre (Okaka and Odhiambo, 2019[32]).

Inadequate infrastructure and low governance capacity increase climate vulnerability

The socio-economic and physical dynamics that shape intermediary cities render them disproportionally vulnerable to climate change. One factor affecting intermediary cities in emerging regions is the rapid expansion of both population and built-up land. In terms of the vulnerability of cities to climate change, Birkmann (2016[33]) notes that the rate of population growth matters more than the size of cities. Fast urban growth tends to outpace the capacity of local authorities to provide public goods and services to city dwellers (Birkmann, 2016[33]). Inadequate urban planning and land management in rapidly expanding intermediary cities also contributes to their vulnerability. Rapid expansion without proper planning leads to sprawl and peri-urbanisation characterised by high levels of inequality and spatial segregation, a prevalence of informal settlements with low or no access to basic services, and environmental degradation (Roberts et al., 2016[34]). This disproportionally exposes intermediary cities to the higher mortality rates associated with weather shocks and to loss of income and livelihoods, while weakening the ability of local authorities to recover from climate disasters.

The frequently weak institutional capacity of intermediary cities further contributes to their vulnerability to climate change. Inadequate decentralisation is one of the main constraints affecting local governments’ institutional capacities. Paterson et al. (2017[35]) argue that the large shift of responsibilities from central governments towards local authorities that is entailed in decentralisation has often not been complemented with adequate financial and human capacity, so that relevant decision-making power remains highly centralised at the national level. Local governments are left with insufficient political power, as well as limited financial and human capacities to manage growing urban demands (Paterson et al., 2017[35]). This is the case among most small and intermediary cities across emerging regions, where local governments have a very limited base for revenue collection and central governments tend to be reluctant to fund urban development effectively, often for political motives (Satterthwaite, 2016[36]). For instance, in Portland, Jamaica, despite decentralised institutional architecture, the power, expertise and resources for Disaster Risk Management (DRM) still remain concentrated at the national level (Blackburn, 2014[37]). Similarly, in Karonga, Malawi, despite a constitutional requirement for decentralised governance, the absence of local governance since 2005 has caused weak urban planning and adaptation mechanisms for managing climate risks (Manda and Wanda, 2017[38]). In contrast, Argentina, which is one of the most decentralised countries in Latin America (OECD/UCLG, 2016[39]), has had some of the most successful climate actions at city level (see the case of Rosario and Santa Fe, in Chapter 4). However, it is important to highlight that decentralisation in Argentina has been a long and complex process (Chapter 4).

Weak institutional capacities also affect the capacity of intermediary cities to manage and implement climate adaptation and mitigation plans. This creates a risk of trade-offs, with other services that are under the responsibility of local authorities, such as education, health and infrastructure, competing with climate adaptation and mitigation costs for financing and attention.

Low economic development characterising many intermediary cities in developing countries translates into low resilience to climate change. Limited diversification and low income per capita not only increase the vulnerability of urban areas to climate threats but are also key factors for the adaptation and response capacity of cities (Garschagen and Romero-Lankao, 2013[40]). Intermediary cities usually lack the economic and political capital of large cities. Many intermediary cities have lower market access, less economic diversification, less ability to attract private investment and lower employment opportunities than larger cities. Most of the economic activities in intermediary cities, especially in low-income countries, depend on primary sectors, such as agriculture, mining and, to some extent, tourism and light manufacturing (Roberts et al., 2016[34]). These factors affect the resilience of intermediary cities and impact local governments’ ability to mobilise resources to cope with and respond to damage resulting from climate-induced disasters (Birkmann, 2016[33]). Additionally, a large share of urban dwellers in intermediary cities are employed in the low-end informal economy, such as construction, street vending, waste management/pickers (Satterthwaite et al., 2020[41]). In this context, climate-induced risks such as flooding, and heat waves can lead to disproportionate loss of livelihood as well as reducing productivity and income.

Unplanned urban expansion with extensive informal settlements increases the vulnerability of inhabitants of intermediary cities to climate risks. High population density and economic activities drive urban land costs upwards, pushing poor urban dwellers to settle on illegal and/or unregulated land. Such settlements have limited safety controls and do not comply with land-use regulations and building standards that can help reduce climate shocks (Satterthwaite et al., 2007[42]). In 2020, approximately 1 billion urban dwellers were living in informal settlements, most of them located in developing countries (Satterthwaite et al., 2020[41]). In other words, these settlements account for one-third of the global urban population (Abunyewah, Gajendran and Maund, 2018[43]).

Urban dwellers in informal settlements are exposed to climate risks due to a series of factors. First, many informal settlements in cities (including in small and medium-sized cities) tend to be located in low-lying and/or river-bank areas that are prone to flooding (Abunyewah, Gajendran and Maund, 2018[43]; John, 2020[44]). Second, most informal settlements lack adequate infrastructure and services, and their poor-quality housing cannot withstand floods, storms and heat waves (Satterthwaite et al., 2020[41]). Third, informal settlements tend to be densely populated with households that are vulnerable due to low economic capabilities (Abunyewah, Gajendran and Maund, 2018[43]). The combination of these factors leaves informal-settlement dwellers ill prepared to cope with the risks brought by climate change. Furthermore, across high-density informal settlements, the line between industrial and residential land use is often blurry. This can cause major safety hazards, for instance through exposure to polluted water and air. This is exacerbated by the fact that these areas often lack space for emergency evacuation, while overcrowded housing complexes can intensify the spread of diseases (Satterthwaite et al., 2007[42]).

High levels of informality and urban inequality also create dual energy-use systems. Urban dwellers in informal settlements, with poor access to infrastructure, energy and electricity, rely on traditional wood-based biomass, which can have negative health impacts in addition to leading to higher particulate matter emissions. This further exacerbates social inequalities (IPCC, 2014[4]) and can put additional strain on climate mitigation efforts.

Climate shocks in rural areas create a ripple effect on intermediary cities

Intermediary cities’ role as a link between rural and urban areas exposes them to the compounded effects of climate change. This section provides a brief explanation of how the intermediation role characterising these agglomerations increases the risk of climate shocks. (Due to the complexity of these processes, interested readers are invited to turn to Chapter 3, which expands this topic and provides a detailed analysis of the channels connecting climate change, rural areas and intermediary cities.)

Intermediary cities are key actors in the rural-urban interface, especially in developing countries. The rural-urban interface encompasses rural areas, small towns and intermediary cities of various sizes, and is an important area that accounts for a large share of both the global population (Berdegué et al, 2014[45]) and smallholder farmers (FAO, 2017[46]). Intermediary cities act as nodes that connect rural, urban and metropolitan areas. They play a large role in helping to reduce rural and urban poverty by acting as hubs for the provision of goods and services and for access to local and international markets, and by facilitating circular migration and enabling the diversification of rural and urban income and livelihoods (Berdegué et al, 2014[45])

Intermediary cities are highly reliant on surrounding rural areas. They rely on the supply of agricultural goods, rural labour and rural consumers for their economic development and the adequate functioning of food systems (Hussein and Suttie, 2016[47]). Similarly, rural areas are highly dependent on intermediary cities for accessing basic public goods and services, circular migration and livelihood diversification (Christiaensen and Todo, 2013[48]; Turok, 2018[49]). For instance, Tacoli and Vorley (2016[50]) highlight the importance of urban areas for rural development across mountain areas of Tanzania, where an increasing number of farmers are switching the production of cash crops to perishable agricultural goods for surrounding urban markets. This study also shows that rural farmers rely on urban areas for diversifying their incomes through activities such as wage earning and contract farming (Tacoli and Vorley, 2016[50]). As such, negative shocks associated with climate change have major implications for the social and economic networks that link urban and rural territories.

Rural livelihoods tend to be particularly threatened by climate variations. Climate change directly affects the incomes and assets of rural households, especially smallholder producers and farmers, who depend on rain for agricultural production and have lower adaptive capacities. Climate change also depletes the assets of rural households and their ability to invest and build adaptive capacities (Dasgupta et al., 2014[51]; FAO, 2016[52]). Moreover, reduced access to urban services in nearby towns and intermediary cities disrupts the economic systems that connect rural and urban areas.

The effects of global warming risk disrupting the channels linking intermediary cities and surrounding rural areas, and in particular food systems and internal migration. Climate change is increasingly disrupting agricultural supply chains, harming infrastructure, livelihoods, remittances and demand for rural products. In parallel, climate change can lead to changes in the pattern and scale of rural-to-urban migration. The following sections give a brief overview of the intricate ways in which climate change can negatively affect food systems, change the scale of migration and create a ripple effect on intermediary cities.

Climate change poses multiple risks to food systems across the rural-urban interface

Food systems are highly vulnerable to disruptions caused by climate change. A food system is the set of activities, actors and interactions that take place along food value chains, from supply of production inputs, agricultural crops and livestock to all post-production activities including processing, transportation, wholesaling and preparation for consumption and disposal. Food systems also include policy and regulatory frameworks around the food economy (IFPRI, 2020[53]). Climate change will affect value chains across food systems, including multiple elements of food production (FAO, 2016[52]).

Climate shocks affecting agricultural production are particularly disruptive for food systems. Evidence increasingly shows that agricultural productivity is reduced by the negative impact of climate change on water resources, soil quality and rural infrastructure (Nyahunda, Tirivangasi and Tirivangasi, 2019[54]), with changes in precipitation and temperature patterns expected to reduce crop yields. Water stress will be the main challenge, with lower soil fertility increasing the impact. Indeed, soil degradation is one of the main channels through which climate change will affect agricultural production. Variations in temperature and moisture, as well as increasing CO₂ levels, are affecting soil and its fertility (Pareek, 2017[55]). Higher temperatures in arid and semi-arid regions will lead to soil salinisation due to a loss in underground water resources (Maharjan and Joshi, 2012[56]). Other regions will be confronted with soil erosion resulting from excessive precipitation and floods, with reduced soil quality and fertility leading to decreasing crop yields (OECD, 2014[57]). Estimates show that global yields of wheat and maize fell by 5.5% and 3.8%, respectively, between 1980 and 2008, compared to a stable climate scenario (Lobell et al., 2011[58]).

Climate change will also affect non-food crops and can lead to significant economic losses in rural and urban areas. Cash crops such as tea, coffee, and cocoa account for a significant share of agricultural production. They are the main source of income for millions of small producers across Africa, Asia and Latin America who are already experiencing yield losses and are forced to carry out agricultural diversification (Dasgupta et al., 2014[51]). In Uganda, for example, climate change is threatening coffee production with extinction within the next 30 to 70 years (Parker et al., 2019[59]). Similarly, in Nicaragua, one of the poorest countries in Central America, cash crops including coffee, maize and beans are particularly vulnerable to climate change (Parker et al., 2019[59]). Rising temperatures will have varying implications on Nicaragua’s agricultural production. For example, 68% of the total land area used for bean production is vulnerable to temperatures rising above 25°C by 2030. Further, rising temperatures may force farmers to switch production to new types of crops that may be vulnerable to changing precipitation and the spread of diseases (World Bank, CIAT and CGIAR, 2015[60]).

Climate shocks are causing damage and losses across food post-production networks while depleting the livelihoods and assets of rural and urban dwellers engaged in the sector. Post-production activities, including storage, processing, transportation and retailing of agricultural goods, are vulnerable to changes in climate. Extreme events can disrupt electricity supply and damage air-conditioning systems, reducing the life span of perishable goods. Similarly, networks of transportation and roads that are essential for the distribution of rural and agricultural goods can be interrupted and damaged by extreme weather events (Vermeulen, Campbell and Ingram, 2012[61]). Disruptions across food systems and networks can cause food insecurity, especially among the most vulnerable urban and rural populations.

Intermediary cities are at the core of the complex networks of food systems and risk bearing the effects of climate-induced disruptions in food supply chains across the rural-urban interface. Their close linkages with and reliance on rural areas imply that negative climate shocks affecting rural livelihoods will affect intermediary cities, too. First, intermediary cities in developing countries tend to rely on primary sectors (Berdegué et al, 2014[45]), with climate effects on rural economies potentially leading to significant economic losses in intermediary cities. Second, losses in rural production can increase the risk of food insecurity across these cities (Reardon and Zilberman, 2018[62]). Third, climate-induced disruptions in the rural-urban interface can limit the capacity of urban dwellers to diversify their livelihoods and incomes across rural and urban areas. Indeed, as will be highlighted in Chapter 3, a large share of the inhabitants of small and medium-sized cities conduct part of their livelihood in rural areas as a means of diversifying their income. Climate-induced disruptions and limitations on rural-urban mobility can put these urban dwellers under strain.

Climate change is also disrupting internal migration

The impact of climate change on rural livelihoods is shifting internal migration patterns between rural and urban areas. As will be discussed in Chapter 3, internal migration remains an important driver of urbanisation for many developing regions, and it is a key factor in income diversification strategies, technology diffusion and rural development (Tacoli, C. et al., 2014[63]; Hussein and Suttie, 2016[47]). As certain climate shocks intensify, migration patterns within the rural-urban interface will be disrupted. The latest (conservative) estimates by the World Bank indicate that by 2050 there will be 216 million climate-induced internal migrants across Sub-Saharan Africa, North Africa, South Asia, East Asia and the Pacific, Eastern Europe and Central Asia, and Latin America (Clement et al., 2021[64]). And many developing countries are already being challenged by climate-induced displacement. For instance, tropical cyclone Idai, which hit Mozambique in 2019, caused the displacement of approximately 146 000 people (Podesta, 2019[65]).

Climate change will particularly affect internal migration in regions with high socio-economic vulnerabilities. In Brazil, for instance, internal migration rates are expected to rise by 9.7% between 2041 and 2070 under a low-emission scenario (Oliveira and Pereda, 2020[66]), while in Bangladesh, climate change will increase the number of internal migrants by 3-10 million by 2050 (Hassani-Mahmooei and Parris, 2012[67]). Moreover, climate-induced migration can increase socio-economic and spatial inequalities. For instance, Brazil’s Northeast region may lose up to 2.5% of its population;7 a significant share of this migration flow will move towards the richer southeast regions of the country (Oliveira and Pereda, 2020[66]).

However, the link between climate change and internal migration is a complex and widely debated issue. Although climate-induced migration has received increasing attention from governments and academics, large knowledge gaps remain, primarily due to: high uncertainty about the effects of climate change in these territories; limited empirical evidence on migration rates and patterns (e.g. rural-to-urban and seasonal); and the role of non-climatic or environmental socio-economic factors in shaping migratory decisions (Tacoli, 2011[68]).

Patterns and causes of rural-to-urban migration are highly complex and context specific. Areas that experience sudden climate shocks, such as floods or hurricanes, can experience high out-migration rates as people are forced evacuate and eventually leave. However, in the case of slow climatic changes, it is harder to establish a direct link to the decision to migrate or not. In Ghana, for example, land degradation and declining crop yields have been key determinants for internal migration, yet poverty and low income are also push factors for internal migration (Van der Geest, 2011[69]). Other socio-economic dynamics in developing countries – such as the urbanisation rate, the country’s economic base (whether it is industrialised or agriculture-based) and its economic growth rate – have a major role in determining migrants’ destinations. For instance, highly urbanised countries in Latin America and the Caribbean, and regions with high economic growth and industrial expansion, tend to have larger rural-to-urban migration flows, while low-income agricultural countries are often characterised by rural-to-rural migration (Tacoli, 2011[68]). In the case of Africa, Henderson et al. (2017[70]) find that climate-driven urbanisation primarily takes place towards towns and cities with an industrial base that can absorb rural labour.

Intermediary cities play an important role in the coping strategies and livelihood diversification plans of rural dwellers. As the effects of climate change intensify many intermediary cities could be confronted with higher migration flows, even though they have limited resources and capacity to meet the needs of their current inhabitants. Rural migrants make up a disproportionate share of the urban poor and tend to live in vulnerable conditions, including working in informal settings, with low income and limited access to public services (Tacoli, C. et al., 2015[71]). As such, they could be further exposed to the climate threats that affect intermediary cities. There are different ways in which local authorities can reduce the vulnerability of rural migrants, such as effective urban planning and increased financing and investment in infrastructure and local government capacity (Tacoli, C. et al., 2015[71]).

How does climate change affect intermediary cities?

The increasing concentration of GHG has a direct effect on global temperatures. Temperatures have been rising each decade since the 1980s (IPCC, 2014[72]). Figure 2.7 shows global temperature anomalies between 1880 and 2019. Each anomaly represents the difference between the temperature registered in a given month and the average monthly temperatures between 1901 and 2000. Since 1980, practically all anomalies have been positive, i.e. global temperatures have been higher than the 20^th century average. Moreover, anomalies after the 1970s show a positive trend, reaching an increase of up to 1.1°C by 2019. In fact, 2019 was the second warmest year ever recorded, after 2016.

Figure 2.7. Global land and ocean temperature anomalies

Rising global temperatures will have a series of detrimental effects on global ecosystems. In 2018, the IPCC’s Special Report on Global Warming warned that exceeding 1.5°C above pre-industrial levels by 2100 would lead to high uncertainty on possible climate scenarios and large risks to humanity. Box 2.3 and Box 2.4 highlight some of the most critical effects of climate change on livelihoods and human well-being.

Despite increasing awareness about climate change, current efforts do not seem enough to limit the rise in global temperatures to 1.5°C before the end of the century. Five years after the Paris Agreement, global GHG emissions continue to increase. The IPCC’s Sixth Assessment brings forward the notion of “code red for humanity” as the planet reaches the brink of a tipping point that will lead to irreversible changes (IPCC, 2021[1]). The report highlights that unless deep measures to reduce CO₂ and other GHG emissions are taken now, the global temperature is projected to exceed the 1.5°C threshold by the early 2040s, i.e. ten years ahead of the projections of the IPCC special report published in 2018. The report notes that the atmospheric concentration of CO₂ is higher than it has been in at least 2 million years; that the global warming experienced between 1970 and 2020 is the fastest in the last 2000 years; and that sea-level rise has been faster since 1900 than in any period in the last 3000 years.

It is highly likely that human activities are the main contributors to global warming. Human activities account for approximately 1.0°C of the increase in temperatures since 1850-1900. The rise in temperatures has contributed to: a) more frequent and extreme land heat waves; b) doubling the number of marine heat waves; c) heavy precipitation; d) ecological and agricultural droughts; and e) increased sea-level rise, among other changes (IPCC, 2021[1]).

Global temperatures are expected to continue rising during the 21^st century. The IPCC projects global warming based on five emissions scenarios for three periods of the 21^st century: near-term (2021-40), mid-term (2040-60), and long term (2081-2100). Average surface temperatures from 1850-1900 are used for comparison for each of these periods. Table 2.1 shows projected global temperatures based on the five scenarios. Across all the scenarios, global surface temperatures are projected to rise, at least until the mid‑21^st century. Moreover, in most of the scenarios, global temperatures are expected to exceed the 1.5°C threshold in the medium and long term. These estimates project higher long-term temperatures than the previous assessment, carried out in 2014 (IPCC, 2021[1]).

Emissions scenarios	Global temperature in the near term (2021-2040)	Mid term (2041-2060)	Long term (2081-2100)
Emissions scenarios	Best estimates	Best estimates	Best estimates
Very low emission pathways (SSP1-1.9)	1.5 ° C	1.6 ° C	1.4° C
Very high emissions pathway (SSP5-8.5)	1.6 ° C	2.4 ° C	4.4° C
Lower (relatively) emissions (SSP3-7.0)	1.5 ° C	2.1 ° C	3.6° C
Emissions with strong climate mitigation (SSP2-4.5)	1.5 ° C	2.0 ° C	2.7° C
Emission with limiting warming below 2° C (SSP1-2.6)	1.5 ° C	1.7 ° C	1.8° C

Note: SSP1-1.9: very low emissions pathways with warming below 1.5°C in 2100 and limited temperature rise during 21^st century. SSP5-8.5: very high emissions pathways with very high warming; SSP3-7.10: lower emissions than SSP5-8.5 with CO₂ emissions doubling by 2100 compared to current levels; SSP2-4.5 and SSP1-2.6: emissions with stronger climate mitigation and lower GHG emissions; SSP1-2.6 scenario to limit warming to below 2°C.

Source: IPCC (2021[1]).

Compared to previous estimates, these new scenarios suggest a higher risk of disruptive climate events as well as a higher likelihood of irreversible changes across different ecosystems. The risks include:

increased and greater regional variability in global precipitation and associated flooding. Global extreme daily precipitation is projected to intensify by 7% for each increase of 1°C, with increased intensity of tropical cyclones. Precipitation will increase particularly in high-latitude areas, whereas regions located in the subtropics will experience declining precipitation. For instance, heavy precipitation and flooding are projected to further intensify in the equatorial Pacific islands, parts of the monsoon regions, as well as in high latitude regions such as in regions of North America and Europe.
increased global mean sea levels throughout the 21^st century. Rising sea levels are projected to affect approximately two-thirds of the global coastline, while the global mean rise will reach two meters by 2100 and five meters by 2150 (under the very high GHG emissions scenario). Sea-level rise will lead to a higher frequency and intensity of flooding, especially in low-lying areas. Extreme climate shocks driven by sea-level rise (i.e. flooding) will occur at least annually in more than half of all tide-gauge locations by 2100 (IPCC, 2021). A rise in ocean temperatures is one of the main factors contributing to sea-level rise. At least 83% of the ocean surface will experience temperature rise throughout the 21^st century across all five scenarios, with a projected increase of 2.89°C under high emissions scenarios.
less efficient CO₂ absorption by ocean and land carbon sinks. As emissions increase, land and ocean carbon sinks are projected to absorb an increasing amount of CO₂ emissions. Over time, under the high GHG emissions scenario, there will be a decrease in their efficiency, which will lead to a higher proportion of CO₂ in the atmosphere. Under an intermediate emissions scenario, ocean and land carbon sinks will also gradually decrease their storing capacity by the mid-21^st century. Under a very low emission scenario, ocean and land carbon sinks will absorb less carbon but this will be due to the declining presence of CO₂ in the atmosphere.

Box 2.3. How climate change is affecting livelihoods and human well-being

Extreme events are disrupting our livelihoods and negatively impacting economic development. Over the next decades, heat waves are expected to increase in terms of both intensity and frequency, while there will be less occurrence of extreme cold weather (IPCC, 2014[72]). These extreme weather events cause large human and economic losses. Between 2015 and 2019, heat waves were the most frequent and deadliest hazards (WMO, 2019[74]). Similarly, cyclones, storms and floods have caused significant economic losses during the last years. For instance, Hurricane Harvey, which struck Texas and Louisiana in 2017, led to an economic loss of more than USD 125 billion (WMO, 2019[74]).

Climate change is projected to cause large economic losses

The world will face significant economic losses unless substantial mitigation actions are taken to meet the 2050 net-zero emissions target set by Paris Agreement. Recent estimates by the Swiss Re Institute (2021[75]) project that under current global temperature trajectories and stated mitigation pledges,1 global temperatures would reach 2.0-2.6°C by 2050, and global GDP will decline by 11%-14%, compared to a scenario of no climate change. Even meeting the Paris Target by mid-century would lead to GDP loss of 4.2%. Under a more severe temperature increase scenario of 3.2°C by 2050, global GDP is set to decline by 18% (Swiss Re Institute, 2021[75]). The economic losses will affect all regions but will be most severe in developing countries in Latin America, Asia and Africa. For instance, a 2°C increase in temperatures by 2050 would lead to a loss in GDP of 11% in Latin America, 14% in the Middle East and Africa, and 15% in Asia, compared to a GDP loss of 7.6% in OECD countries (Swiss Re Institute, 2021[75]). The IPCC’s special report on 1.5°C also warned of global economic losses, although with lower estimates. According to the IPCC, a global temperature rise to 3.6°C by 2100 under a no-mitigation scenario will lead to a total global GDP loss of 2.6%, compared with a GDP loss of 0.3% at 1.5°C and 0.5% at 2°C (Hoegh-Guldberg et al., 2018[24]).

Countries and regions with higher levels of social vulnerability are predisposed to higher losses. Baarsch et al. (2020[76]) estimate that African countries have already sustained economic losses of 10%-15% of GDP due to climate change between 1986 and 2015. Countries in East and West Africa suffered the highest impact due to their high reliance on the agriculture sector and limited adaptation systems, the study reports. The countries least affected were those with more diversified economies, a higher natural resource endowment or a larger share of the service sector. Indeed, climate shocks seem to have deeper effects on countries highly reliant on agriculture following a productivity loss. Declining yields have a major impact on food prices, affecting the well-being of rural and urban households as well as the terms of trade at country and global level (Lobell et al., 2011[58]; Knox et al., 2012[77]; Sultan and Gaetani, 2016[78]; Hertel et al., 2010[79]). Countries that are net food exporters will benefit from higher commodity prices, while the terms of trade of net importer countries will deteriorate. Similarly, households that are net agricultural producers would tend to benefit from higher prices, while poverty among net consumer households will worsen (Hertel et al., 2010[79]).

Climate change imposes large economic costs in various ways. It causes physical damage to infrastructure and transportation services and increases the demand for public health services (IPCC, 2014[72]). Climate change also disrupts production factors, such as capital stock (e.g. factories and production centres) and productivity, including the capacity of factories to operate, as well as the use of roads and other infrastructure networks. Climate shocks directly affect certain sectors, such as agriculture, and indirectly affect less climate-sensitive sectors such as manufacturing (Lecocq and Zmarak, 2007[80]). For instance, damage caused by sea-level rise is expected to reduce Viet Nam’s economic growth prospects by 2050, mainly due to the degradation of infrastructure and agricultural yields (Arndt et al, 2015[81]).

Climate change will also lead to losses in labour productivity and working hours. Heat stress reduces labour productivity, as it negatively affects workers’ physical and cognitive capacity. According to the ILO (2019[82]), provided that global temperatures do not rise by more than 1.5°C by 2100, there will be a 2.2% loss in total working hours by 2030. This corresponds to a total global loss of 80 million full-time jobs and estimated economic losses of USD 2 400 billion by 2030. The subregions of Western Africa and Southern Asia will be particularly impacted by heat-stress-induced economic losses, with 43 million and 9 million full-time job losses, respectively, by 2030 (ILO, 2019[82]). A rise in temperatures will especially affect outdoor and physically difficult labour. Knittel et al. (2020[83]) find that within Europe, countries in the Mediterranean region, such as Italy, Malta and Spain, will be the most impacted by a reduction in work ability or labour productivity. Other regions, such as Southeast Asia, India and oil exporting countries, will also experience a severe reduction in labour productivity (Knittel et al., 2020[83]).

1. Compared to a world with 0°C temperature change.

Box 2.4. Climate change will cause severe health risks and increase in mortality

Climate change presents a series of health risks, especially to vulnerable populations in low-income countries. Extreme weather events such as heat waves, droughts, increased air and water pollution, and changes in precipitation pose direct and indirect health risks (Berry et al., 2018[84]). Watts et al. (2018[85]) find that intense heat waves affected 125 million adults between 2000 and 2016. Indirect effects take place through changes in ecosystems caused by climate change (Watts et al., 2015[86]). These include increased air and water pollution, which can lead to an increase in the spread of vector-borne diseases, such as dengue. Since 1950, the spread of vector-borne diseases has increased by 9.4% (Watts et al., 2018[85]). In 2019 the number of reported cases of dengue in Central and Latin America alone reached 2.8 million, with 1 250 deaths in the region. Between August and October 2019, countries including Brazil, Mexico, Colombia, Nicaragua, Philippines, Thailand and Malaysia accounted for 85% of the total 1.05 million dengue cases reported (WMO, 2019[74]). Other indirect effects include health risks caused by changes in crop nutrient values, food security and respiratory diseases. Health risks caused by water scarcity are on the rise, despite not being accounted for in current estimates. Moreover, health risks caused by climate change vary based on socio-economic conditions, social factors and social norms such as gender, etc. (Watts et al., 2015[86]).

Extreme events are increasing mortality rates around the world. Between 1995 and 2015, extreme temperatures were the cause of 27% of total weather-related deaths; 90% of them were caused by heat waves, while almost 92% of heat-related deaths took place in high-income countries, especially in Europe (UNISDR, 2016[87]). According to WHO estimations (2014[88])1, under a base-case scenario climate change is expected to cause 250 000 additional deaths per year between 2030 and 2050. The study finds that the highest mortality rates by 2030 will be in Sub-Saharan Africa, while by 2050 South Asia will be the most affected region (WHO, 2014[88]). However, estimates on climate-induced mortality might be underestimating death rates, especially when other indirect causes, such as poverty, hunger, malnutrition and climate-induced conflicts are taken into account (Parncutt, 2019[89]).

Although climate change is a global phenomenon, emerging and developing countries are disproportionally vulnerable to climate risks. For instance, Revi et al. (2014[21]) note that low- and middle-income countries accounted for 95% of deaths recorded from floods and storms between 2000 and 2013 (Revi et al, 2014[21]). Asia is one the regions most affected by weather-related disasters. Between 1995 and 2015, Asia accounted for 2 495 of the total 6 457 weather-related disasters recorded globally, which affected 3.7 billion people and killed 332 000 persons (UNISDR, 2016[87]). Additionally, Southeast, Southern and Eastern Asia accounted for 89% of the total 2.74 billion people killed and affected worldwide due to climate-related disasters between 2000 and 2012 (Busby et al., 2018[22])

The effects of climate change on health will put significant pressure on public health sectors, which are already under strain, especially in developing countries. The health implications of climate change come with large financing costs. WHO estimates that annual health care costs associated with climate change will be USD 2-4 billion between 2030 and 2050, and that developing countries with poor health infrastructure will face larger challenges in coping with rising costs (WHO, 2018[90]). Health-care cost estimations vary highly by region as well as the type of climate variabilities and health risks taken into account. The OECD (2016[91]) estimates that the rise in the global health-care cost of outdoor air pollution will range from USD 21 billion to USD 176 billion between 2015 and 2060, while the adaptation cost for malaria and diarrhoea will reach USD 2 billion between 2010 and 2050. A large share of the latter costs will be primarily borne by countries in sub-Saharan Africa (Pandey, 2010[92]).

1. The estimations consider various climate-induced causes of death, including heat waves, coastal flooding, vector-borne diseases including diarrhoea (especially among children under the age of 15), malaria, dengue, as well as undernutrition and other related causes. The death rates are expected to vary highly based on future economic growth rates, underlying health conditions, socio-economic dynamics as well as public health systems and universal health coverage. However, the results do not capture climate-induced indirect causes of death.

Heat, drought, floods and rising seas will directly affect cities

The following sections highlight some of the most critical effects of climate change on urban areas. They analyse potential differences among cities of different sizes when it comes to climate shocks and outline key socio-economic characteristics of intermediary cities that shape their vulnerability to climate change.

Cities will be exposed to rising temperatures and more and hotter heat waves

Cities have experienced a higher frequency of heat waves since the mid-20^th century, and this trend is expected to increase in the next decades. The built-up land of cities plays a key role in this process, since it absorbs heat, which intensifies both the duration and intensity of heat waves (KFW, 2015[93]). In London, for example, annual number of nocturnal heat waves have increased by four days since the 1950s, with an average increase of heat intensity of 0.1°C (Revi et al, 2014[21]). By 2050, London’s nocturnal heat-wave intensity in August is expected to increase by 0.5°C, with a 40% rise in the frequency of heat-wave episodes (Revi et al, 2014[21]; OECD, 2010[27]). Overall, by 2050, heat waves will affect around 970 cities, with a disproportionate number of cities exposed in Asia, Africa and North America (UCCRN, 2018[26]). By 2090, the number of people exposed to heat waves in Eastern African cities, including Kampala (Uganda), Lusaka (Zambia), Blantyre-Limbe (Malawi) and Likasi, Kolwazi and Lubumbashi (DRC) will increase two thousand fold (Rohat et al., 2019[31]).

Extreme heat waves can cause mortality due to heat-related deaths, famine and the spread of infectious and non-infectious diseases. For example, the heat wave in Europe in 2003 claimed 70 000 victims, of whom a significant proportion came from urban areas (OECD, 2010[27]).

Urban heat islands (UHI) make cities particularly vulnerable to increasing temperatures (Rosenzweig et al., 2015[94]). The replacement of natural land cover with pavement and buildings makes cities particularly vulnerable to increasing temperatures. As cities grow, and population and infrastructure density increase, the surrounding environment is altered. This directly affects the capacity of a city to react to the consequences of climate change (Revi et al, 2014[21]). Growing urban areas tend to experience higher temperatures than surrounding places with a lower density (rural areas) and other agglomerations that are less congested (Chapman et al., 2017[95]). The effect of UHI is compounded by cities’ economic and industrial activities, as well as by the use of motor vehicles, which produce heat and increase urban temperatures (Ningrum, 2017[96]; World Bank, 2010[97]). The fact that urban centres are made of heat-absorbing materials, such as concrete, with less vegetation and green areas than rural areas, reduces their capacity for natural water retention and their infiltration potential (KFW, 2015[93]).

The combination of UHI and climate change also contributes to air pollution and GHG emissions. Air pollution is one of the most prominent issues facing urban areas across the world. UHI and warmer temperatures increase the concentration of air pollutants, including ozone, acid aerosol and small particulates (OECD, 2010[27]). The fact that emissions in cities have higher concentrations of pollutant particles leads to higher urban pollution intensity (Li et al., 2018[98]). Interactions between UHI and urban air pollution can increase the effects of both on urban areas. Warmer temperatures caused by UHI can disperse and increase the mixing of air pollutants and lead to higher concentrations of pollutants (like PM 10), eventually leading to an increase in longwave radiation (Li et al., 2018[99]). At the same time, an increase in aerosol pollution enhances urban heat islands (Cao et al., 2016[100]).

It is important to note that not just large cities are affected by UHI. As discussed in Box 2.5, intermediary cities in Brazil are already experiencing the effects of UHI despite having limited industrial activity.

Box 2.5. Urban heat islands in medium-sized and small cities in Brazil

While research on UHI is predominantly focused on large cities, the growth of small and medium-sized cities implies that they will also be impacted. Assessing UHI in these urban centres is important due to their rapid population growth and the expansion of built-up land, which implies a reduction in vegetation and absorption of heat into the ground. UHI can present a larger challenge to small and medium-sized cities as they tend to be poorly planned.

A study of three such cities in Brazil – Paranavaí (Paraná), Rancharia (São Paulo) and Presidente Prudente (São Paulo) – highlights changes in nocturnal heat islands. UHI have been detected in the three cities despite their low population density (Paranavaí, 87 316; Rancharia, 25 828; Presidente Prudente, 223 749) and limited industrial activities. The cities, which have different urban forms, are characterised by high temperatures during spring and summer, and milder winds and lower temperatures in winter. UHI were detected to cause increases in temperatures of 2.3°C to 5°C in Paranavaí and 3.5°C to 6°C in Presidente Prudente. UHI in these cities were mainly underpinned by changes in urban land cover and morphology. Areas with higher population density were also associated with a higher intensity of UHI.

Source: Cardoso et al (2017[101]).

Episodes of excessive heat in cities are becoming more common in developing countries. The heat index developed by the US National Weather Service measures the perception of heat when temperature and humidity are combined. When there is a heat index of at least 40.6º C (105º F), people face a danger of heat exhaustion and heat strokes with prolonged exposure and physical activity. Throughout this section, such events are referred as “extreme heat events”.

On average, extreme heat events are more frequent in developing countries than in richer economies. In 2016, cities in high-income countries went through an average of four events of excessive heat, while cities in poorer countries had an average of ten events. In developing countries, cities that are characterised by high temperatures8 are experiencing longer and more frequent extreme heat events. In 2016, cities characterised by low temperatures went through an average of five extreme heat events with an average duration of 5.3 days, while in warmer cities there were on average 12 extreme heat events that lasted 8.8 days. Moreover, the frequency of such events is increasing faster in warmer cities (Figure 2.8). From 1987 to 2016, the frequency of these events grew annually by 0.3% in low-temperature cities and the duration by 0.5%; in high-temperature cities, the frequency grew by 0.7% and the duration by 0.6%.

Figure 2.8. Average frequency and duration of extreme heat events (1987-2016)

There is an inverse relationship between rising temperatures and GDP, with growing heat stress resulting in a loss of jobs and productivity (ILO, 2019[82]). Garcia-Leon et al. (2021[103]) estimate that economic losses triggered by heat waves in Europe in the years of 2003, 2010, 2015 and 2018 amount to 0.3-0.5% of European GDP, compared to the average loss in GDP between 1981 and 2010. The losses have been larger in heat-prone areas and places where work takes place outdoors. In a study of US counties, Deryugina and Hsiang (2014[104]) find that income per capita declines as temperatures increase: a day with an average temperature of 29ºC decreases annual income by 0.065% compared to a day with an average temperature of 15ºC. Evidence also suggests that higher temperatures have a stronger impact on developing countries. Dell et al. (2012[105]) find that, on average, a 1ºC rise in temperatures in a given year reduces economic growth by 1.3 percentage points, but only in poor countries.

This relationship between GDP and temperatures is also observed across cities of developing countries. Figure 2.9, and the corresponding tables in Annex 2.A1, show the results of an analysis exploring the connection between the frequency of extreme heat events and GDP on cities of different sizes, and differentiates between cities with high and low average temperatures. Results worth highlighting include:

There is a negative relationship between the number of extreme heat events and GDP at city level across time (after controlling for city characteristics such as population and build-up). However, the strength of this relationship is more acute in cities with low temperatures, i.e. the effect is significantly larger in cities characterised by low temperatures, independently of their size. For instance, a 1% increase in the frequency of these events is associated with a decrease of 0.37% in the GDP of low-temperature cities of 50 000 to 100 000 inhabitants, while for high-temperature cities of the same size the effect is -0.10%. This difference can be seen across cities of all sizes.
The largest effect of extreme heat events is observed in cities that are both large and characterised by low temperatures. In cities with low temperatures and more than 1 million inhabitants, a 1% increase in the frequency of extreme events is associated, on average, with a decrease of 0.65% in GDP. From 1987 to 2016, the frequency of extreme heat events in low-temperature cities with more than 1 million inhabitants grew annually by 0.1%, which would translate into an average annual drop in GDP of 0.065%.
In high-temperature cities, the negative effect of extreme heat events on GDP is observed only in cities with fewer than 500 000 inhabitants. Moreover, the impact is more acute in cities with a population of 50 000 to 100 000 than in cities with a population of 100 000 to 500 000. In the former, a 1% increase in the frequency of extreme heat events is associated with a 0.1% drop in GDP, and in the latter with a drop of 0.06%. Given that the frequency of these events in these cities has grown annually by around 0.68%, this would translate into an annual drop in GDP of 0.07% and 0.04%, respectively.

The lower impact on GDP of extreme heat events in high-temperature places can be due to adaptation to heat. First, people in warm places seem to show a higher tolerance to extreme heat. Singh et al. (2018[106]) use Regional Internet Search Frequencies for air conditioning in India as an indicator of thermal discomfort, and find that people living in places with high average temperatures have a higher tolerance of heat. This adaptation or tolerance can be physiological, behavioural of psychological. Second, the fact that extreme heat events are more common in high-temperature places will increase the likelihood of these cities taking active measures to mitigate heat. A review of different articles about heat adaptation shows that, in low- and middle-income countries, the most common extreme heat responses are behavioural or cultural adaptations by individuals and communities, with little institutional involvement (Turek-Hankins et al., 2021[107]). Other factors can be linked to the fact that economic activities taking place in high-temperature cities are already intended for hot weather. For instance, outdoor activities might not take place in the warmest hours of the day, and people performing indoor activities might have ventilation devices already available, such as fans. In high-temperature cities, the larger effect in those with fewer than 500 000 inhabitants can be partly explained by their tighter relationship to rural areas and the primary sector. Agriculture accounted for 83% of global working hours lost to heat stress in 1995, and it is estimated to have accounted for 60% in 2013 (ILO, 2019[82]). This follows from the fact that agricultural work is mostly outdoors and requires a high physical effort that is impaired under heat stress.

Extreme heat events are a threat to a city’s economy and the well-being of its citizens regardless of the baseline climate, and the frequency of these events has been growing faster in high-temperature places. In order to offset the consequences of extreme heat events, cities need to take adaptation measures that protect their citizens and economic activities as well as those of surrounding areas. However, adaptation strategies should be planned carefully, as many initiatives that prioritise immediate and short-term climate risk reduction can reduce the opportunity for transformational adaptation. Moreover, maladaptation can lock in the vulnerability of the people that are intended to be protected, while being difficult and costly to change once implemented (Dodman et al., 2022[108]).

Figure 2.9. Effects of extreme heat events on the GDP of cities in developing countries

Water stress and droughts will be more frequent in cities of developing countries

Droughts are extreme events in which precipitation levels, soil moisture and groundwater fall below normal for several months or even years. The intensity, frequency and land coverage of droughts are influenced by global climatic patterns, as well as by local factors. For this reason, it is difficult to quantify the extent to which climate change has influenced this phenomenon. However, increasing evidence suggests that, during the last 50 years, anthropogenic emissions have contributed to more frequent warm events. Rising global temperatures since the 1970s have altered atmospheric circulation patterns and contributed to higher evaporation of land moisture, surface water and plant soils, which in turn have intensified droughts overall (Dai, 2011[109]). Moreover, dry soils and diminished plant cover can further reduce rainfall in a dry area, creating positive feedback and making droughts more persistent (C2ES, 2021[110]).

Droughts are increasingly affecting water scarcity in urban areas. Since 2000, 79 of the world’s largest urban areas have experienced urban droughts. Urban areas located in arid, semi-arid and humid regions are more susceptible to droughts. Climate change not only exacerbates the frequency and extent of droughts, but is also making it harder for cities to cope with water scarcity since it tends to deplete commonly used water reservoirs such as dams (Zhang et al., 2019[111]). At the same time, while cities are experiencing increasing droughts and higher temperatures, they also face higher demand for water due to rising urbanisation and increasing economic activities (Flörke, Schneider and McDonald, 2018[112]). Global water demand is expected to increase by up to 55% by 2050 (OECD, 2012[113]).

Rising global temperatures combined with fast and unplanned urban growth risk increasing the exposure of cities to water stress and droughts. Growth in urban population implies an increasing demand for water and very often the depletion of natural resources. This will impose increasing strains on local governments seeking to manage declining freshwater resources. Furthermore, urban sprawl exposes vulnerable dwellers to limited access to water or polluted water sources (Gebre and Gebremedhin, 2019[114]). Water stress is already occurring in fast-growing small and medium-sized cities in Nepal and India. In Nepal, the cities of Dharan and Dhulikhel have been experiencing fast population growth, and this coupled with the increasing effects of climate change and urban sprawl has depleted ground water resources (The Straits Times, 2019[115]). In India, the city of Chennai experienced drought with severe water shortages in 2019, leading it to import 2.5 million litres of water (The Straits Times, 2019[115]).9

Changes in precipitation patterns are also expected to affect groundwater reserves and recharge rates, putting additional pressure on urban dwellers. Persistent droughts and low precipitation can lead to higher depletion of groundwater reserves due to an increase in demand as well as the extraction of groundwater for irrigation systems, especially in low-income countries (OECD, 2014[57]). In regions like Africa, a decline in groundwater supply can affect 90 million people living in rural areas, while in Asia, changes in water supply are expected to impact food security affecting 60 million people (Dasgupta et al., 2014[51]). Box 2.6 highlights the increasing strains on water resources caused by droughts and increasing water demand in South Africa’s Eastern Cape Province.

Box 2.6. The effects of drought on cities in South Africa’s Eastern Cape Province

South Africa’s Eastern Cape Provinces has a network of intermediary cities and small towns and is also one of the country’s regions most vulnerable to drought. The province encompasses medium-sized cities such as Port Elizabeth, Nelson Mandela Bay, Buffalo City and East London (John, 2012[116]). It has been experiencing severe droughts since 2015. In 2019 it was declared a drought-disaster region following severe water shortages across both rural and urban areas (Mahlalela et al., 2020[117]). At the start of 2020, the region continued to face worsening drought conditions, with the area’s supply dam at only 6% of its capacity (SABC, 2020[118]). Several urban areas in the region almost ran out of piped water (Mahlalela et al., 2020[117]). Cities in the region such as Port Alfred have faced prolonged water shortages, leading an increasing number of residents to rely on water tanks (SABC, 2020[118]).

Prolonged and intense droughts in Eastern Cape region are causing a series of negative socio-economic impacts. Droughts combined with increasing urbanisation and rising demand for water are affecting the well-being of people in the province (PMG, 2018[119]). The region’s principal sources of livelihood, livestock and communal farming, are particularly vulnerable to droughts. A study conducted in 2014 in the region’s O.R. Tambo district found that farmers lost or delayed the sale of their livestock due to droughts and often ended up with lower prices. The droughts also caused psychological stress among farmers, higher dependence on government grants, poverty and an overall lack of security (Muyambo, Jordaan and Bahta, 2017[120]).

Water stress and droughts present socio-economic and welfare losses, especially for urban dwellers. By 2050, more than 650 million urban dwellers living across 500 cities will face declining availability of fresh water, with a projected reduction in streamflow of more than 10% compared to current levels (UCCRN and C40, 2018[121]). Current estimates suggest that one in four cities faces water stress and that this trend will continue to increase (McDonald et al., 2014[122]).

Droughts and water stress cause health and economic risks, especially in cities of low and middle-income countries. They can cause public health issues such as diarrhoeal diseases, especially among children, and can lead to the faster spread of diseases, especially in densely populated areas with low water and sanitation services (Ashraf et al, 2016[123]). Droughts can also cause food insecurity (Kareem et al., 2020[124]), and are associated with economic and welfare loss. A study of 78 large Latin American cities between 2005 and 2014 concluded that droughts lead to a decline in employment and number of working hours, as well as negatively impacting informal employment (Desbureaux and Rodella, 2019[125]). The economic losses are attributed in particular to the negative effects of droughts on electricity supply, as hydropower is the main source of electricity in Latin America (Desbureaux and Rodella, 2019[125]).

Cities of different sizes face different exposure to flooding and storm surges

Flooding is one of the most prominent climate shocks affecting urban areas. Climate change is associated with urban flooding through three main channels: a) floods caused by increasing sea levels; b) floods from overflowing rivers and glacial melting; and c) floods caused by prolonged precipitation or storm surges, which result from an abnormal rise of water generated by a storm affecting coastal areas (Satterthwaite et al., 2007[42]). Flooding poses serious risks to a significant proportion of the urban population worldwide. In particular, floods are expected to increase in Asia, Africa and Latin America (UN-HABITAT, 2011[9]).

A larger share of the population in intermediary cities than in large cities is potentially exposed to riverine floods and storm surges. Figure 2.10 shows the share of cities potentially exposed to storm surges and riverine floods (left), and the share of population potentially exposed to storm surges and built-up areas potentially exposed to riverine floods (right), by city size in 2015. The figure shows that the share of cities potentially exposed to storms or floods increases with city size – but that the share of population and built-up area potentially exposed decreases with city size. For instance, only 4% of the cities in the smallest group were potentially exposed to storm surges, while almost 90% of their population was potentially exposed to this type of climate shock. In contrast, 17% of the cities in the largest group were potentially exposed to storm surges, which concerned 41% of their population. A similar pattern is observed in the case of riverine flooding. In cities with 50 000 to 100 000 inhabitants, 26% of the population and 60% of their built-up areas were potentially exposed to riverine flooding, while in cities with more than 1 million inhabitants, almost half of the population but only 38% of their built-up areas were potentially exposed to riverine floods.

Figure 2.10. Exposure to riverine floods and storm surges, 2015

As previously discussed, the vulnerability of a city is partly explained by its growing process. In many cities at potential risk of flooding and storms, the built-up area has expanded quickly, leading to a decrease in urban density and a reduction of green areas. This dynamic has previously been found to exacerbate the risks of floods, as was the case with Hurricane Harvey in Houston (Zhang et al., 2018[126]). Moreover, a reduction of green space increases the imperviousness of urban areas (i.e. the ability of urban areas to absorb water), which increases storm-water runoff and damages the urban water-cycle system (Salvadore et al., 2015[127]).

Poor urban planning, inadequate management of land use and lack of effective governance render urban areas susceptible to flooding. Geography and socio-economic factors characterising certain urban areas increase the risk of flooding: proximity to coastal areas, deforestation, concrete, poor-quality infrastructure and housing, and lack of an adequate drainage system (Satterthwaite et al., 2007[42]; Campbell-Lendrum and Corvalán, 2007[128]). Roads and concrete infrastructure prevent the absorption of water from rainfall into the ground. Inadequate drainage paired with poor waste management usually result in clogging, which prevents excess water from being absorbed into the drainage. This is exacerbated by the fact that, in many cities, drainage systems are not built within the urban fabric and the drainage is sometimes obstructed by buildings (Satterthwaite et al., 2007[42]). Box 2.7 uses the example of the low-lying city of Can Tho, Viet Nam, to illustrate how flooding can strain local water resources.

Box 2.7. Flooding in Can Tho, Viet Nam

Can Tho is the fourth largest city in Viet Nam, with around 1.4 million inhabitants. It is located in the Mekong Delta region on the southern bank of the Hau River (Huynh et al, 2020[129]). The city is particularly vulnerable to flooding, typhoons, storms and droughts, and the frequency of flooding has increased in recent years. Between 1900 and 1960, the region experienced four severe floods, while there were 11 between 1961 and 2011. Can Tho city alone experienced severe flooding events in 2011 and 2013. The exposure to extreme flooding is caused by a rise in sea level as well as land subsidence caused by groundwater exploitation (Quang Vinh Ky, 2018[130]).

The rise in flooding in Can Tho has severe implications, especially for the poorest and most vulnerable populations. It has created large economic losses in the agriculture and energy sectors as well as damaging water resources (Huynh et al, 2020[129]). Flooding caused by sea-level rise is particularly straining the city’s water resources by increasing the salinity of freshwater resources, leading to a severe shortage of water to meet growing urban demand (Rajesh, 2018[131]).

Floods cause socio-economic losses and infrastructure damage. They can destroy energy and electricity networks, damage and contaminate water-storage services and disrupt transportation and road services (UN-HABITAT, 2011[9]). Most importantly, floods can raise health risks by increasing the likelihood of malaria, dengue and other water-borne diseases. For instance, high humidity, which increases with high temperatures and precipitation, contributes to the spread of the Aedes mosquito and dengue fever. Flooding induced by climate change is therefore expected to lead to an increase in dengue-endemic areas, both globally and in some developing regions, including China. The health problems associated with floods are often compounded by poor urban waste management, which usually results in waste getting into drinkable water (Campbell-Lendrum and Corvalán, 2007[128]). Floods in cities can be an important cause of death. To cite two examples, floods in Mumbai in 2005 caused more than 1 000 deaths, while more than 900 people died in severe flooding in Algiers in 2001 (Satterthwaite et al., 2007[42]).

Coastal cities face a major threat from rising seas

Rising sea levels present high risks to populations living in coastal cities. According to the IPCC’s Sixth Assessment, the mean global sea level is expected to rise throughout the 21^st century, increasing by 0.28‑0.55 meters by 2100 under the very low GHG emissions pathway (IPCC, 2021[1]).10 Estimates suggest that sea-level rise will affect 824 million people by 2030, and that this number will reach 1.2 billion by 2060 (Church et al., 2013[132]).

Sea level rise is one of the main consequences of increasing ocean heat and rising land temperatures. Global sea levels have been rising since the 20^th century, mainly due to thermal expansion but also due to the fast melting of ice sheets in Greenland and Antarctica. These two factors have accounted for 75% of sea-level rise since 1971 (Cubasch et al., 2013[133]). Between 1901 and 1990, the global mean rate of sea-level rise was approximately 1.4 mm per year; between 1970 and 2015, the rate increased to 3.2 mm per year; and during the period 2006-15, it reached 3.6 mm per year (Oppenheimer et al., 2019[134]).

Global sea-level rise particularly threatens low-lying coastal areas and small islands. Low coastal areas in developing countries are particularly vulnerable due to fast population growth and built-up expansion, which have accelerated the anthropogenic subsidence of these areas. Anthropogenic subsidence is expected to outpace other sources of sea-level rise mentioned above.

Under both the 1.5°C and 2°C climate scenarios, and if no further adaptation plans are implemented, 136 coastal cities of more than 1 million inhabitants are at risk of flooding. Many of these urban centres are located in South and Southeast Asia (Hoegh-Guldberg et al., 2018[24]). Jevrejeva et al. (2018[135]) estimate that a rise of 2°C by 2040 means that more than 90% of global coastal lines will experience sea-level rise of above 0.2 m. According to their findings, people living in large coastal cities such as Lagos, Guangzhou and New York are particularly at risk and face limited time to establish adaptation measures.

A large share of the world’s urban population lives in close proximity to coastal areas. In 2015, around 22% of city dwellers in developing countries lived in areas below 50 meters of altitude and within 5 kilometres of the coast.11 This represents more than 635 million people. The share of cities with more than 1 million inhabitants located near the coast is much higher than the share of smaller cities. One-quarter of the big cities of the developing world are located in low-lying coastal areas, with 36% of the population, or 450 million people, residing in these cities (Figure 2.11). A much lower percentage of cities with fewer than 1 million inhabitants is located in these areas, but a sizeable share of the population is still at risk. For instance, 10% of cities with 100 000 to 500 000 inhabitants are located in low-lying coastal areas, and they count 11% of the population of cities of this size, or 107 million people.

Figure 2.11. Share of population living near coastal areas, by city size (2015)

Globally, more than 570 cities, counting more than 800 million inhabitants, will be exposed to sea-level rise of 0.5 meters by 2050 (UCCRN and C40, 2018[121]). Rising seas will particularly affect Asian cities, as well as cities in deltaic areas (Church et al., 2013[132]). By 2050, 80% of the population affected by sea-level rise will be in Southeast and East Asia; among the most threatened countries in the region are China, India, Bangladesh, Indonesia and Viet Nam (Revi et al, 2014[21]). In Africa, sea-level rise is expected to reach 0.38m by 2080, and the average number of people affected will increase from 1 million per year in 1990 to 25 million by 2050, and potentially 70 million by 2080 (Douglas et al., 2008[136]). Coastal cities with low drainage infrastructure, such as Lagos (Nigeria), Mombasa (Kenya), Mumbai (India) and Dhaka (Bangladesh), will face major challenges in managing floods caused by sea-level rise (Revi et al, 2014[21]).

Low-lying coastal cities play a key role in Asia’s urbanisation process and are a key element of the region’s export-oriented economy and overall global trade. Increasing urban sprawl and population growth across coastal areas will further exacerbate the regions’ vulnerabilities if adequate adaptation strategies are not implemented. Urbanisation causes land subsidence, especially in areas that are already below sea level, such as the Pearl and Mekong river deltas. Some of Asia’s largest urban areas, including Bangkok, are already sinking at 4 cm per year, while in Jakarta land is subsiding at 6 cm per year (Fuchs, Conran and Louis, 2011[137]). In Asian cities such Khulna in Bangladesh, sea-level rise is already causing large challenges such loss of assets, damage and contamination of water and sanitation infrastructure (Box 2.8) (KFW, 2015[93]).

Box 2.8. Sea-level rise in Khulna, Bangladesh

Khulna, the third largest city in Bangladesh, is highly vulnerable to sea-level rise. With 663 000 inhabitants, Khulna is one of the country’s most important economic centres (Roy et al., 2018[138]) and it hosts a large number of industries, including chemicals, food processing and packaging, and shipbuilding. Its vulnerability to rising seas is due to its geographic position – in the Ganges river delta, inland from the Bay of Bengal – as well as to rapid population growth and high levels of poverty and inequality. The city’s infrastructure and water resources are under strain, and GHG emissions are increasing (Zermoglio et al., 2020[139]).

Khulna is vulnerable to a series of climate-induced threats, such as rising temperatures, changes in precipitation and flooding. The effects of sea-level rise and other climate-induced threats are particularly damaging as the city lacks effective urban planning and its dwellers face acute shortages of safe housing, infrastructure, safe drinking water, sewerage and other solid waste management systems (Roy et al., 2018[138]). Sea-level rise is putting additional strains on the city, as causing clogging in the draining systems and leading to contamination of water resources and the spread of water-borne diseases. Sea-level rise is also increasing the salinity levels of the city’s groundwater sources, reducing freshwater availability for the growing number of urban dwellers (Zermoglio et al., 2020[139]).

Small and medium-sized cities in coastal areas may benefit economically from their geographic position, but they are increasingly exposed to sea-level rise. As highlighted by Roberts (2014[140]), coastal small and medium-sized cities tend to be economically dynamic, with diversified industries and sought-after housing markets. However, these cities are also disproportionately exposed to storm surges caused by sea-level rise, and tend to be more heavily polluted than inland cities. Because these cities are growing quickly, rising seas could create additional challenges for protecting vulnerable populations living in informal settlements (Roberts, 2014[140]). For example, Beira, Mozambique’s second largest city, is highly vulnerable to rising sea levels. In 2019, the city was hit by Cyclone Idai, causing 1 000 deaths and large losses in assets and infrastructure (Williams, 2021[141]).

How are intermediary cities contributing to climate change?

Cities play a key role in national and international efforts to mitigate climate change. Urban areas, which cover merely 2% of the earth’s surface, account for 60% of global GHG emissions (UN-Climate Action, 2018[142]), and 70% of total CO₂ emissions (IEA, 2021[143]). The high GHG emissions of cities stem from their high energy demand: indeed, urban areas account for 75% of global energy use (IEA, 2021[143]). This is expected to increase as urban demand for energy and infrastructure continues to grow. By 2050, energy demand from cities is expected to grow by 70% compared to 2013 levels, accounting for 66% of total global energy demand and increasing CO₂ emissions from the energy sector by 50% (IEA, 2016[144]).

Cities contribute to GHG emissions mainly through three gases: carbon dioxide (CO₂), methane (CH₄) and nitrous oxide (N₂O). These gases are the result of energy conversion, landfills and urban solid waste, as well as land conversion (from rural to urban areas). The emissions are also the by-product of energy production for meeting urban domestic demand, for manufacturing construction materials used in urban infrastructure and for the provision of food to city dwellers. Moreover, these emissions interact and may even reinforce each other. For instance, the higher concentration of carbon monoxide (CO) in cities, which is produced by transportation emissions, traps emissions such as CH₄ and prolongs their presence in urban areas (OECD, 2010[27]). Cities also produce additional gases that are specific to economic activities taking place in urban areas, such as ozone (O₃), one of the main GHGs and the third most important pollutant after CO₂ and CH₄, as well as sulphur hexafluoride (SF₆), which is emitted during the production of refrigerants and semiconductors (OECD, 2010[27]).

However, CO₂ remains the key driver of GHG emissions worldwide. Approximately 72% of the gases contributing to climate change in 2018 were fossil-fuel-related, with CO₂emissions coming mainly from coal, oil, and natural gas combustion (Olivier and Peters, 2020[145]). The top emitters of CO₂ (and other GHGs) are China (which accounts for 30% of global CO₂), the United States (14%), the European Union (9%), India (6.9%), Russia (4.6%) and Japan (3.2%). Developing countries account for only 33% of CO₂ emissions, but they bear some of the most severe consequences of climate change. Overall, the potential for reducing GHG emissions is higher in non-OECD countries – where urban planning is either lacking or at early stages across a large number of urban areas – than in OECD countries (IEA, 2016[144]).

Cities in developing countries are becoming important contributors to GHG emissions

Emissions are rapidly accelerating in developing countries. OECD countries were the main contributors of CO₂ before the 1970s following an energy-intensive industrialisation process. But over the past 30 years, a large share of emissions has come from developing countries, especially in Asia. Since 2010, emissions from Asian countries account for the largest share (IPCC, 2014[72]). Figure 2.12 compares CO₂ emissions per capita in 1970 against their annual growth rates between 1970 and 2017 for different geographical regions. It shows that OECD countries in America and Europe had the highest levels of CO₂ emissions per capita in 1970, at 16.2 and 8.1 tonnes per person, respectively; in contrast, China emitted only 0.9 tonnes per person. The growth trajectories of emissions in these regions have since reversed: from 1970 to 2017, OECD countries in Europe and America experienced negative growth rates, while China’s CO₂ emissions per capita expanded rapidly, with an average annual growth rate of 4.3%. Likewise, Asia and the Middle East started from a low level but caught up fast, with growth rates above 3% during the period.

Figure 2.12. Growth rates of CO₂ emissions per capita

Emerging regions are also shifting the global landscape of energy demand. Countries in Asia account today for two-thirds of the growth in global energy demand, with demand currently growing twice as fast in Southeast Asia as in China. At country level, India accounts for 30% of the growth in energy demand in the world, and by 2040 will account for 11% of the increase in global demand for energy (OECD/IEA, 2017[147]). As for China, in 2018 it had the highest growth in energy demand since 2012, and this accounted for one-third of total growth in global energy demand, according to the IEA (2018[146]).

As the economies of developing countries transform, the GHG emissions from their cities will continue to increase. Many developing countries have industrialised, with some becoming centres for global manufacturing. This process, coupled with fast urbanisation rates and growing wealth, has led to more emissions from urban areas in emerging economies. By 2030, emissions from cities in non-OECD countries will account for 81% of global energy use (OECD, 2010[27]).

Income differences within cities also affect emission rates. Affluent urban dwellers have a higher emissions rate due to higher energy consumption, use of private transport and consumption of goods characterised by highly embedded carbon. In India, for example, the annual CO₂ emissions of the richest 1% of the population are four times those of the poorest 38% (UN-HABITAT, 2011, p. 51[9]). Moreover, the livelihoods of many poor urban households depend on recycling and reuse of waste, which can yield negative emission levels. Although recycling produces emissions, negative emissions arise when the emissions “saved” exceed the emissions produced (Satterthwaite, 2009[148]).

Big cities produce most urban CO₂ emissions in developing countries. In 2015, cities of more than 1 million inhabitants produced around 13 million tons of CO₂ on average. This is almost five times more than the average amount of CO₂ produced by cities with a population of 500 000 to 1 million, 37 times more than cities of 100 000 to 500 000, and 132 times more than cities of 50 000 to 100 000.

Per capita CO₂ emissions are also lower in intermediary cities than in large cities. Figure 2.13 compares CO₂ per capita emissions in 2000 and 2015. It shows that, in 2015, an average person in an intermediary city produced 2.3 kilo tonnes of CO₂, compared to 3.7 kilo tonnes per person in large cities. It also shows that CO₂ emissions per capita increased over the period, with per capita emissions in intermediary cities rising by 2015 to the level in large cities in 2000.

Figure 2.13. CO₂ emissions per capita in intermediary and big cities, 2000 vs. 2015

Large cities also account for the largest share of CO₂emissions in most economic sectors. Excluding high-income countries, cities with more than 1 million inhabitants accounted for 55% of all CO₂ emissions in 2015 (Figure 2.14). However, as the figure shows, the contribution to CO₂ emissions by city size depends on the sector of activity: large cities produce close to two-thirds of CO₂ emissions in the transport and industry sectors, and more than half of the emissions in the residential sector, while intermediary cities account for the largest share of emissions in the energy and agriculture sectors.

Figure 2.14. Contribution of cities of different sizes to CO₂ emissions

There are, however, regional variations in the share of CO₂ emissions by city size. Although big cities represent a small percentage of urban centres, they are responsible for the lion’s share of urban CO₂ emissions. Figure 2.15 shows that only 5% of the cities in Latin America and the Caribbean have more than 1 million inhabitants, but they account for 68% of the region’s urban CO₂ emissions, while small and medium-sized cities account for the rest (32%). In Sub-Saharan Africa, big cities represent 2% of urban centres but account for 62% of urban CO₂ emissions. In contrast, intermediary cities in some regions emit more CO₂ than big cities. This is the case of India, where intermediary cities (98% of all Indian cities) accounted for 57.5% of CO₂ emissions in 2015.

Figure 2.15. Percentage of cities with more than 1 million inhabitants and % of CO₂ emissions

The contribution to CO₂ emissions by different sectors is fairly similar across intermediary cities.12 Overall, the industry and energy sectors are the largest contributors to CO₂ emissions. Figure 2.16 shows that in big cities, the industrial sector emits the largest share, while in intermediary cities energy production contributes the most to CO₂ emissions. The transport and residential sectors account for a smaller but still relevant share of emissions across all city sizes. Agriculture is the smallest contributor to urban emissions. This does not mean that agriculture is not an important source of overall CO₂ emissions, but rather that is not such a relevant factor in urban CO₂ emissions.

There are important differences in sectoral emissions across regions. In China and India, the energy sector accounts for the highest share of emissions, particularly in intermediary cities. In other regions, such as Western Asia and Southeast Asia, the energy sector is not as relevant, but it is higher in intermediary cities. Latin America has the highest share of industrial and transport emissions, which are fairly similar across city sizes. Sub-Saharan Africa is a particular case as residential emissions account for a large share of CO₂ emissions, reaching 55% of total CO₂ emissions in cities with a population of 50 000 to 100 000.

Figure 2.16. Composition of emissions in cities of different sizes

CO₂ emissions are evolving differently across cities of different sizes

Urban emissions of CO₂have grown in all regions. From 2000 to 2015, CO₂ emissions from cities outside high-income countries increased by 4.6% on average. They grew across all regions, particularly in China (7.3%) and India (4.1%), while the lowest increase was in Sub-Saharan Africa (1.7%). Although CO₂ emissions grew more rapidly in big cities on average, regions followed different trends. In Western Asia, for instance, the smallest cities saw a 6.2% increase in emissions, while big cities’ emissions grew by 2.4%.

Overall, CO₂ emissions from the energy sector grew fastest during the period 2000-15. This sector grew on average by 5.5%, followed by the industrial sector (4.9%) and transport (4.9%), while the residential sector (1.3%) and agriculture (1.6%) experienced the lowest growth rates.

Sectoral CO₂ emissions followed different trends depending on the size of the city. CO₂ emissions from the energy and transport sectors grew faster in smaller cities, reaching 6.6% and 5.1% respectively in cities of 50 000-100 000 and 100 000-500 000 inhabitants. CO₂ emissions from the residential, agriculture and industrial sectors grew more rapidly in larger cities (Table 2.2). As such, the energy sector contributed most to the growth of CO₂ emissions in cities with fewer than 1 million inhabitants and was responsible for 58% to 62% of the increase. In cities with more than 1 million inhabitants, the industrial sector was the main contributor to CO₂ emissions growth, with up to 46% of the increase between 2000 and 2015.

	Energy		Residential		Industrial		Transport		Agriculture
City size	Rate	Contribution	Rate	Contribution	Rate	Contribution	Rate	Contribution	Rate	Contribution
50K- 100K	6.6%	61.7%	0.6%	2.7%	4.1%	27.4%	5.1%	7.4%	1.5%	0.8%
100K-500K	5.1%	58.0%	0.6%	2.3%	4.0%	31.6%	5.0%	7.5%	1.6%	0.6%
500K-1M	5.1%	58.6%	0.9%	2.3%	3.9%	32.5%	4.7%	6.3%	1.8%	0.3%
>1M	5.7%	39.6%	1.9%	4.5%	5.5%	46.1%	4.9%	9.5%	1.7%	0.2%

Note: “Rate” refers to the annual average growth rate between 2000 and 2015, while “contribution” refers to the contribution to emissions growth by each sector during the same period. The sample does not include cities in high-income countries.

Source: Author’s own calculations using GHS (2015) data.

Cities of more than 500 000 inhabitants tend to have a higher CO₂ intensity. The increase of CO₂ emissions in cities has historically been tied to economic development, since important industries relied on the burning of fossil fuels. To understand this relationship, the carbon intensity of GDP measures how much CO₂ is produced per dollar of GDP, or the carbon cost of economic development (Wang, Yang and Qi, 2020[149]). In developing countries, the CO₂ intensity of GDP decreased between 2000 and 2015, especially in intermediary cities. As a result, in 2015, CO₂ intensity was lower in cities with fewer than 500 000 inhabitants than in bigger cities.

Cities in developing economies have yet to break the link between economic growth and CO₂ emissions. A decrease of CO₂ intensity is known as decoupling. Decoupling refers to breaking the link between “environmental bads” (CO₂ emissions) and “economic goods” (GDP growth), and it can either be absolute or relative (OECD, 2002[150]). Absolute decoupling is usually described as the way to make economic growth environmentally sustainable, and it happens when countries manage to decrease their CO₂ emissions while their economy continues to grow. On average, however, cities in developing countries are undergoing relative decoupling, as their CO₂ emissions continue to grow, but at a much lower rate than GDP (Figure 2.17). This can be seen as a gain in efficiency, but it does not remove the link between economic growth and environmental impact (Ward et al., 2016[151]).

Figure 2.17. Growth rate of CO₂ intensity and its components

GDP, population and urban expansion contribute to growing CO₂ emissions in cities

CO₂ emission levels are strongly associated with the degree of urbanisation. As cities grow and their wealth increases, CO₂ emissions tend to be higher. Figure 2.18 shows the relationship between GDP and CO₂ levels in 2015 for the sample of cities outside high-income countries. It shows a positive trend that follows an S-shape. This shape seems to result from the relationship between GDP and CO₂ among different city groups, i.e. the strength of this relationship seems to grow with city size, only to eventually decrease among large cities with more than 1 million inhabitants. Figure 2.19 also shows the relationship between CO₂ emissions and both population and built-up land for the same sample of cities in 2015. The data suggest an overall linear relationship between these variables (expressed in logarithmic terms), and also suggest that the relationship between CO₂ and these variables is different among city groups.

Figure 2.18. CO₂ emission, GDP, population and built-up land across cities of different sizes

Changes in GDP, population, and built-up land area can be expected to affect the level of CO₂ differently, depending on city size.13 Figure 2.19 shows regression estimates for the expected change in CO₂ emissions resulting from changes in the GDP, population and built-up land in the sample of cities outside high-income countries (Table 2.A.4 in Annex 2.A1 shows detailed results of these regressions). Three main points emerge. First, an increase in GDP is associated with higher CO₂ emissions, and this effect increases with city size.14 In cities with more than 500 000 inhabitants, the effect of GDP on emissions is significantly larger than in smaller agglomerations. On average, an increase of 1% in GDP among cities with more than 1 million inhabitants is associated with a 0.38% increase in CO₂ emissions, while the rise in emissions is 0.13% among cities of 50 000 to 100 000. Second, higher levels of population are associated with higher CO₂ emissions in cities with less than 500 000 inhabitants. A 1% increase in population is associated with a 0.15% and a 0.22% increase in total CO₂ emissions among cities of 50 000 to 100 000 and 100 000 to 500 000 inhabitants, respectively. Third, changes in built-up area have a significant effect on CO₂ emissions only in cities of fewer than 500 000 inhabitants. In such cities, a 1% increase in built-up land is associated with an increase of 0.18% in CO₂ emissions in cities of 50 000 to100 000 inhabitants and an increase of 0.15% in cities of 100 000 to 500 000 inhabitants.

Figure 2.19. Expected change in CO₂ emissions from changes in GDP, population, and built-up land

The extent to which changes in production and urbanisation affect CO₂ emissions also depend on the activity of a city’s main economic sectors. In Annex 2.A1, Figure 2.A.1 shows the results from a series of econometric models focusing on emissions in the residential, transport and industry sectors. (Columns 2, 3 and 4 of Table 2.A.4 in Annex 2.A1 display the results of the regressions). Some general trends emerge from this analysis:

Larger cities tend to have a stronger relationship between GDP and CO₂ in the transport and industrial sectors. For cities of more than 1 million inhabitants, a 1% increase in GDP is associated with a 0.5% increase in transport CO₂ emissions; in cities of fewer than 100 000 inhabitants, the increase is close to 0.3%. A similar pattern is found in CO₂ emissions from the industrial sector, where 1% increase in GDP ranges from 0.22% in cities with less than 100 000 inhabitants to 0.3% in cities with more than 1 million people. In contrast, an increase in GDP is associated with a slight decrease in residential CO₂ emissions in cities of fewer than 500 000 inhabitants: a 1% increase in GDP is associated with a decrease in residential emissions of around 0.03%.
Population plays a key role in the increase of emissions in the residential sector among intermediary cities. Indeed, the estimated effect of a 1% increase in population is an increase in emissions of 0.24% in cities with less than 100 000 inhabitants to 0.46% in cities of 100 000 t0 500 000 inhabitants. In contrast, for transport emissions, the relationship between population and emissions is negative: a 1% increase in population is associated with a decrease in emissions of 0.21% to 0.29% across all city sizes.
Built-up expansion is positively associated with higher CO₂ emissions, but only in cities of fewer than 500 000 inhabitants. The effect is particularly acute in the residential sector, where a 1% increase in built-up land is associated with an increase of 0.33% to 0.34% in CO₂ emissions in cities with fewer than 500 000 inhabitants.

These results are in line with literature studying the dynamics between urbanisation and CO₂ emissions. GDP has proved to be a significant driver of CO₂ emissions in cities in developing countries, as the Environmental Kuznet’s Curve (EKC) theory suggests: as cities become richer, emissions grow until they reach a certain level of development, from which point emissions go down. Evidence on the existence of the EKC was found by Castells-Quintana et al. (2020[152]) and He et al. (2017[153]). Previous studies also suggest that built-up expansion drives up CO₂ emissions by converting land for commercial, residential or industrial purposes (Wang et al., 2018[154]; Zhou et al., 2015[155]) and through the carbon lock-in of infrastructure, as long-life capital stocks can lock in CO₂ emissions for very long periods (World Bank, 2010[97]).

Why climate mitigation efforts are particularly important for intermediary cities

Intermediary cities deserve special attention in terms of climate mitigation efforts for several reasons. First, emissions in small and medium-sized cities have grown constantly since 2000, particularly those from the energy and transport sectors. Second, urbanisation trends imply that these cities will continue to grow in the next years, most likely without proper planning and with a high risk of sprawl, and that some of them eventually will become large metropolitan areas. If not adequately planned, these cities risk being locked onto an unsustainable path and will face the challenges big cities face today.

Although emission levels are shaped by the way cities grow, the relationship between urban dynamics and CO₂ growth is far from clear. The multiple studies exploring the effect of a country’s urbanisation15 on CO₂ emissions in most cases find a positive relationship between the two (Parikh and Shukla, 1995[156]; Cole and Neumayer, 2004[157]; Wei, Yagita and Inaba, 2003[158]). Nevertheless, the relationship between the growth of cities and CO₂ emissions is far from simple, and is based on multiple factors. Some city-level studies highlight the role of urban density on CO₂ emissions, finding a relation between increasing population density in cities and decreasing CO₂ emissions per capita (Wang et al, 2020[159]; Castells-Quintana et al., 2015[160]). Other studies link higher density with higher emissions because of greater congestion (Gaigné, Riou and Thisse, 2012[161]). Some studies have focused on the effect of wealth on emissions, finding evidence of a Kuznet’s Curve: wealth and emissions have an inverted U-shape relation (He et al., 2017[153]). The spatial distribution of cities and the number of urban centres has also been identified as a relevant factor. One major conclusion stands out from these studies: There are many factors related to urbanisation that influence CO₂ emissions, and many results are country- and metric-specific (i.e. depends on the measure used). This raises the importance of understanding the dynamics between urban trends and CO₂ emissions among cities in developing economies.

Intermediary cities are key in the achievement of a zero-carbon future that is in line with economic objectives and society’s welfare. According to the Coalition for Urban Transitions (2019[162]), cities with fewer than 750 000 inhabitants have more than half of the carbon abatement potential of large agglomerations around the world. These carbon savings will come from improving commercial and residential buildings (58%) and enhancing the transport sector (21%), materials efficiency (16%) and waste management (5%). The good news is that this abatement will largely be possible through technically feasible measures that are already available in cities (such as decarbonising urban electricity grids, shifts in public transportation or improving the efficiency of heating and cooling systems in buildings). To bring these measures to scale and reach their full abatement potential, local governments will need to work closely with national and regional governments16 and involve the community and protect citizens from the challenges that may arise from the transition. In this way, cities can become compact, connected and clean urban centres that attract investment, promote sustainable economic development by increasing productivity and innovation, and create a healthy and secure environment for their residents.

There is large scope for limiting the effects of urban expansion among intermediary cities in developing countries. A large number of these cities – especially medium-sized cities – face a significant lack of infrastructure. The fact that this gap will eventually be addressed brings an unprecedented opportunity for building low-carbon infrastructure and housing that takes account of climate hazards, facilitates energy efficiency and uses recycled or green materials in construction (Seto et al., 2014[163]; IPCC, 2018[164]). Likewise, many intermediary cities in developing countries lack urban planning and land management systems. The early urbanisation stage in some developing countries, especially in Africa, provides scope for implementing policies that promote the use of pedestrian space and green areas (such as parks), and that push for investment in low-carbon transportation systems, which can help to reduce urban emissions while improving the population’s well-being (IPCC, 2018[164]).

Growth in intermediary cities and increased standards of living will translate into higher GHG emissions unless systemic changes are implemented. As populations, grow there will be higher demand for energy, infrastructure, transportation, housing and other public services. According to the IPCC (2014[4]), with the currently available technology and with the global population estimated to reach 9.3 billion by 2050, addressing the corresponding infrastructure demand will generate approximately 470 gigatonnes (Gt) of CO₂. A large share of the needed infrastructure will be built in intermediary cities in low- and middle-income countries (IPCC, 2014[4]; Seto et al., 2014[163]; Grubler et al., 2012[165]). Indeed, the growth in energy use will be highest in small and medium-sized cities, which are currently characterised by low and moderate final energy use. These cities are expected to increase their energy use by a factor of 6.1 (for fast-growing small cities) and 1.6 (for medium-sized cities), whereas energy use in the largest cities is expected to rise by a factor of 0.5 (Seto et al., 2014[163]; Grubler et al., 2012[165]). That the largest projected growth in energy use is in small cities with fewer than 500 000 inhabitants corresponds to the fact that these cities are expected to undergo the highest population growth, with very high elasticity in growth of energy demand (Seto et al., 2014[163]; Grubler et al., 2012[165]). These cities also have low capacity in the institutional, financial and technical resources needed to manage and mitigate climate threats, rendering them particularly important for targeted implementation strategies (Seto et al., 2014[163]).

Future GHG emissions in intermediary cities will differ highly across regions, based on population size as well as the economic structures of these urban centres. City-level GHG emissions depend on geography, economic structure and specialisation, and energy use (Seto et al., 2014[163]). A significant share of manufacturing industries and production are located in urban areas, which drives their GHG emissions upward (UN-HABITAT, 2011[9]). For example, many Asian intermediary cities have a large industrial base and are the growth engines of their respective regions. In Hai Phong,17 a major industrial city in Viet Nam, CO₂ emissions from energy sources are accounted for 13.2 million tonnes in 2010 and were estimated to increase to 49.6 million tonnes in 2020 (OECD, 2016[166]). Similarly, in the Malaysian cities of Jahor Bahru and Pasir Gudang (located within Isklandar, a special economic zone), CO₂ emissions increased from 5 million tonnes to 18.5 million tonnes between 2000 and 2012 (OECD, 2018, p. 42[167]). As for Africa, Godfrey and Zhao (2015[168]) note that 69 African cities of more than 500 000 inhabitants, from across 35 countries, emitted 240 million tonnes of CO₂ 2012. By 2030, the total emissions of CO₂ are projected to increase by 61%, to 386 million tonnes, for cumulative CO₂ emissions of 6 billion tonnes between 2012 and 2030 (Godfrey and Zhao, 2015[168]).

System innovation for intermediary cities: Towards sustainable policy solutions

A shift in perspective is required to address the systemic challenges brought by climate change and increasingly volatile ecosystems, as well as to achieve the global development goals. This implies thinking of climate actions not as partial measures but as a whole system. It calls for transforming unsustainable systems towards lower energy demand and for shifting from traditional measures of development (mainly GDP) towards improved well-being outcomes (OECD, 2021[12]). This shift in perspective is referred to as a well-being lens. Applying a well-being lens to climate actions can help to catalyse systemic changes that are sustainable by design and are better able to build synergies between improved environmental quality and the attainment of well-being goals. Through this approach, the benefits (and costs) of a specific mitigation measure will be weighed according to their impact on health, education or security, and not just on economic welfare (OECD, 2019[169]).

Improving well-being outcomes is key for building resilience. Indeed, addressing the vulnerabilities of intermediary cities to climate change also implies addressing the underlying socio-economic conditions of these cities. A large share of the population of intermediary cities is often characterised by low socio-economic conditions that heighten their vulnerabilities to climate change. Moreover, intermediary cities tend to be dependent on climate-sensitive sectors (such as agriculture) and to face large gaps in “risk-reducing” infrastructure (such as roads, waste and drainage systems, access to safe water, electricity, etc.) (IPCC, 2014[170]). Urban dwellers in informal settlements are particularly at a disadvantage due to their limited access to these services. Climate vulnerabilities in intermediary cities are largely shaped by the compounded factors of low socio-economic development and low institutional capacities. Climate risks serve as multipliers of threats in these agglomerations. As such, climate actions should take into consideration how multiple economic sectors are affected by climate change, and these actions should aim to strengthen the socio-economic conditions of urban dwellers. They should also tap into some of the opportunities characterising intermediary cities, such as access to education and training opportunities, as well as a strong linkage to bigger markets (UCLG, 2016[171]).

Applying a well-being lens to climate action can also facilitate better mitigation outcomes at the city and national levels. The IPCC’s pathways for net-zero transitions largely depend on the prioritisation of policies. Attaining low energy demand or transformational pathways requires systemic changes that yield improved well-being outcomes. Yet most climate policies today tend to focus on decarbonising systems rather than addressing the systemic changes needed for net-zero transition. These approaches also depend on carbon removal technologies that are not in use yet. In contrast, applying a well-being lens to climate action would imply taking advantage of synergies across different objectives and planning in advance for possible trade-offs. The well-being lens builds on systems thinking, with the core aim of accelerating climate mitigation by redesigning systems that centre on well-being goals (such as health, accessibility, biodiversity, etc.). Indeed, the proposed well-being lens will lead to designing systems that are less reliant on energy and will ultimately lead to lower emissions while simultaneously attaining well-being outcomes (OECD, 2019[169]). The well-being lens also fosters better co‑operation across key stakeholders and enables local governments to prioritise mitigation strategies more effectively, while changing the community’s perception of environmental objectives (OECD, 2019[169]). Climate mitigation strategies are most effective when they are localised and effectively communicated to local stakeholders. This helps to maximise public engagement while also enabling longevity of climate actions.

Applying a well-being lens to local and national policies also calls for a shift from analysing parts to viewing systems as a whole. In other words, it implies building systems that are sustainable by design (OECD, 2019[169]). For instance, adopting systems thinking in food systems can enable policy makers to understand the complexity of these systems, including the distribution of activities across space and actors, and to see how food systems relate to climate change. What is required is a holistic approach that can capture the interconnected economic, ecological, social and spatial dimensions of urban systems.

Applying a well-being lens requires a three-step thinking process that can be applied to various sectors or systems. This includes: envisioning the desired outcome of a functioning and sustainable urban system; understanding the underlying factors and key stakeholders that sustain undesirable results; and changing policy packages, governance systems and budgeting in order to attain the desired outcome (Figure 2.20).

Figure 2.20. The well-being lens process

Conclusion

Intermediary cities play a critical role in addressing climate change, and they deserve a more relevant role in the international agenda. This chapter has highlighted that the growth dynamics of intermediary cities, paired with their underlying socio-economic characteristics, make these cities particularly important for climate action. Moreover, as intermediary cities grow, their GHG emissions will increase. Ignoring this can potentially undermine national and international efforts to curb global emissions. For these reasons, the inclusion of intermediary cities in local, national and international climate policies is critical today. If intermediary cities are endowed with adequate resources and targeted policies, they have the potential to contribute to the reduction of global GHG emissions and to act as a laboratory for innovative solutions. However, this calls for a shift in policy approaches.

References

[43] Abunyewah, M., T. Gajendran and K. Maund (2018), “Profiling Informal Settlements for Disaster Risks”, Procedia Engineering, Vol. 212, https://doi.org/10.1016/j.proeng.2018.01.031.

[81] Arndt et al (2015), “The economic costs of climate change: A multi-sector impact assessment for Vietnam”, Sustainability (Switzerland), Vol. 7/4, pp. 4131-4145, https://doi.org/10.3390/su7044131.

[123] Ashraf et al (2016), “Infrastructure, Incentives, and Institutions”, American Economic Review, Vol. 106/5, pp. 77-82, https://doi.org/10.1257/aer.p20161095.

[76] Baarsch, F. et al. (2020), “The impact of climate change on incomes and convergence in Africa”, World Development, Vol. 126, https://doi.org/10.1016/j.worlddev.2019.104699.

[7] Bae, C. and W. Richardson (2004), Urban Sprawl in Western Europe and the United States, Routledge, California.

[45] Berdegué et al (2014), Inclusive Rural-Urban Linkages, https://rimisp.org/wp-content/files_mf/1431869344123InclusiveRural_UrbanLinkages_edited.pdf (accessed on 11 February 2021).

[84] Berry, P. et al. (2018), “Assessing Health Vulnerabilities and Adaptation to Climate Change: A Review of International Progress”, International Journal of Environmental Research and Public Health, Vol. 15/12, https://doi.org/10.3390/ijerph15122626.

[33] Birkmann (2016), “Boost resilience of small and mid-sized cities”, Nature, Vol. 537/1, pp. 605-609, https://www.nature.com/news/boost-resilience-of-small-and-mid-sized-cities-1.20667.

[37] Blackburn, S. (2014), “The politics of scale and disaster risk governance: Barriers to decentralisation in Portland, Jamaica”, Geoforum, Vol. 52, https://doi.org/10.1016/j.geoforum.2013.12.013.

[18] Brueckner, J. (2000), “Urban Sprawl: Diagnosis and Remedies”, International Regional Science Review, Vol. 23/2, pp. 160-171, https://doi.org/10.1177/016001700761012710.

[10] Burgalassi, D. and T. Luzzati (2015), “Urban spatial structure and environmental emissions: A survey of the literature and some empirical evidence for Italian NUTS 3 regions”, Cities, Vol. 49, pp. 134-148, https://doi.org/10.1016/j.cities.2015.07.008.

[22] Busby, J. et al. (2018), “In harm’s way: Climate security vulnerability in Asia”, World Development, Vol. 112, pp. 88-118, https://doi.org/10.1016/j.worlddev.2018.07.007.

[110] C2ES (2021), Drought and Climate Change, Center For Climate and Energy Solutions, https://www.c2es.org/content/drought-and-climate-change/ (accessed on 11 July 2021).

[29] Cambaza, E. et al. (2019), “Outbreak of Cholera Due to Cyclone Kenneth in Northern Mozambique, 2019”, International Journal of Environmental Research and Public Health, Vol. 16/16, p. 2925, https://doi.org/10.3390/ijerph16162925.

[128] Campbell-Lendrum, D. and C. Corvalán (2007), Climate change and developing-country cities: Implications for environmental health and equity, https://doi.org/10.1007/s11524-007-9170-x.

[100] Cao, C. et al. (2016), “Urban heat islands in China enhanced by haze pollution”, Nature Communications, Vol. 7/1, https://doi.org/10.1038/ncomms12509.

[101] Cardoso, R. et al. (2017), “Assessment of Urban Heat Islands in Small- and Mid-Sized Cities in Brazil”, Climate, Vol. 5/1, https://doi.org/10.3390/cli5010014.

[17] Carruthers, J. and G. Ulfarsson (2002), “Fragmentation and Sprawl: Evidence from Interregional Analysis”, Growth and Change, Vol. 33/3, pp. 312-340, https://doi.org/10.1111/1468-2257.00193.

[160] Castells-Quintana et al. (2015), “Climate change and the geographical and institutional drivers of economic development”, Centre for Climate Change Economics and Policy Working Paper, No. 198, University of Leeds and the London School of Economics and Political Science, Leeds and London, http://www.lse.ac.uk/grantham.

[152] Castells-Quintana, D., E. Dienesch and M. Krause (2020), “Density, Cities and Air Pollution: A Global View”, SSRN Electronic Journal, https://doi.org/10.2139/ssrn.3713325.

[95] Chapman, S. et al. (2017), “The impact of urbanization and climate change on urban temperatures: a systematic review”, Landscape Ecology, Vol. 32/10, pp. 1921-1935, https://doi.org/10.1007/s10980-017-0561-4.

[48] Christiaensen, L. and Y. Todo (2013), Poverty Reduction during the Rural-Urban Transformation The Role of the Missing Middle, http://econ.worldbank.org.

[132] Church et al. (2013), Sea Level Change, IPCC, Cambridge and New York, https://www.ipcc.ch/site/assets/uploads/2018/02/WG1AR5_Chapter13_FINAL.pdf.

[64] Clement, V. et al. (2021), Groundswell: Acting on Internal Climate Migration - Part 2, World Bank, Washington DC, http://www.worldbank.org.

[162] Coalition for Urban Transitions (2019), Climate Emergency: Urban Opportunity, Coalition for Urban Transitions, Washington DC and London, https://urbantransitions.global/en/publication/climate-emergency-urban-opportunity/ (accessed on 13 July 2021).

[157] Cole, M. and E. Neumayer (2004), “Examining the Impact of Demographic Factors on Air Pollution”, Population and Environment, Vol. 26/1, https://doi.org/10.1023/B:POEN.0000039950.85422.eb.

[133] Cubasch et al. (2013), “Introduction”, in Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge (UK), New York (USA).

[109] Dai, A. (2011), “Drought under global warming: a review”, WIREs Climate Change, Vol. 2/1, https://doi.org/10.1002/wcc.81.

[51] Dasgupta, P. et al. (2014), “Rural Areas”, in Edward and Raholiao (ed.), Climate Change 2014: Impacts, Adaptation and Vulnerability, Cambridge University Press, Cambridge, https://www.ipcc.ch/report/ar5/wg2/.

[105] Dell, M., B. Jones and B. Olken (2012), “Temperature Shocks and Economic Growth: Evidence from the Last Half Century”, American Economic Journal: Macroeconomics, Vol. 4/3, pp. 66-95, https://doi.org/10.1257/mac.4.3.66.

[104] Deryugina, T. and S. Hsiang (2014), Does the Environment Still Matter? Daily Temperature and Income in the United States, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w20750.

[125] Desbureaux, S. and A. Rodella (2019), “Drought in the city: The economic impact of water scarcity in Latin American metropolitan areas”, World Development, Vol. 114, pp. 13-27, https://doi.org/10.1016/j.worlddev.2018.09.026.

[108] Dodman et al. (2022), “Cities, Settlements and Key Infrastructure”, in Climate Change 2022: Impacts, Adaptation and Vulnerability, Cambridge University Press, Cambridge.

[136] Douglas, I. et al. (2008), “Unjust waters: Climate change, flooding and the urban poor in Africa”, Environment and Urbanization, Vol. 20/1, pp. 187-205, https://doi.org/10.1177/0956247808089156.

[25] Ebi et al. (2018), The 1.5°C Health Report: Synthesis on health and climate science in the IPCC SR1.5, University of Washington, http://climatetracker.org/ipcc-report-on-health/.

[4] Edenhofer, O. et al. (eds.) (2014), Climate change 2014: Mitigation of Climate Change: Working Group III contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, United Kingdom and New York, USA.

[46] FAO (2017), The State of Food and Agriculture. Leveraging Food Systems for Inclusive Rural Transformation, FAO, Rome, http://www.fao.org/3/a-I7658e.pdf (accessed on 17 February 2021).

[52] FAO (2016), The State of Food and Agriculture 2016: climate change, agriculture and food security, Food and Agriculture Organisation, Rome, http://www.fao.org/publications/sofa/2016/en/.

[20] Filho, W. et al. (2019), “Assessing the impacts of climate change in cities and their adaptive capacity: Towards transformative approaches to climate change adaptation and poverty reduction in urban areas in a set of developing countries”, Science of The Total Environment, Vol. 692, pp. 1175-1190, https://doi.org/10.1016/j.scitotenv.2019.07.227.

[6] Florczyk, A. (2019), GHS Urban Centre Database 2015, multitemporal and multidimensional attributes, http://data.europa.eu/89h/53473144-b88c-44bc-b4a3-4583ed1f547e (accessed on 13 July 2021).

[173] Florczyk, A. (2019), GHS Urban Centre Database 2015, multitemporal and multidimensional attributes, http://data.europa.eu/89h/53473144-b88c-44bc-b4a3-4583ed1f547e (accessed on 13 July 2021).

[112] Flörke, M., C. Schneider and R. McDonald (2018), “Water competition between cities and agriculture driven by climate change and urban growth”, Nature Sustainability, Vol. 1/1, pp. 51-58, https://doi.org/10.1038/s41893-017-0006-8.

[137] Fuchs, R., M. Conran and E. Louis (2011), “Climate Change and Asia’s Coastal Urban Cities”, Environment and Urbanization ASIA, Vol. 2/1, pp. 13-28, https://doi.org/10.1177/097542531000200103.

[161] Gaigné, C., S. Riou and J. Thisse (2012), “Are compact cities environmentally friendly?”, Journal of Urban Economics, Vol. 72/2-3, https://doi.org/10.1016/j.jue.2012.04.001.

[103] García-León, D. et al. (2021), “Current and projected regional economic impacts of heatwaves in Europe”, Nature Communications, Vol. 12/1, https://doi.org/10.1038/s41467-021-26050-z.

[40] Garschagen, M. and P. Romero-Lankao (2013), “Exploring the relationships between urbanization trends and climate change vulnerability”, Climatic Change, Vol. 133/1, https://doi.org/10.1007/s10584-013-0812-6.

[19] Gasper et al (2011), “Social and economic impacts of climate change on the urban environment”, Current Opinion in Environmental Sustainability, Vol. 3/3, pp. 150-157, https://doi.org/10.1016/j.cosust.2010.12.009.

[114] Gebre, T. and B. Gebremedhin (2019), “The mutual benefits of promoting rural-urban interdependence through linked ecosystem services”, Global Ecology and Conservation, Vol. 20, https://doi.org/10.1016/j.gecco.2019.e00707.

[168] Godfrey and Zhao (2015), “The Contribution of African Cities to the Economy and Climate”, New Climate Economy, London, https://newclimateeconomy.report/workingpapers/wp-content/uploads/sites/5/2015/09/NCE-African-Cities-Economy-Climate_technical-note.pdf (accessed on 13 July 2021).

[165] Grubler, A. et al. (2012), “Urban Energy Systems”, in Global Energy Assessment - Toward a Sustainable Future, https://iiasa.ac.at/web/home/research/Flagship-Projects/Global-Energy-Assessment/Chapte18.en.html (accessed on 13 July 2021).

[67] Hassani-Mahmooei, B. and B. Parris (2012), “Climate change and internal migration patterns in Bangladesh: an agent-based model”, Environment and Development Economics, Vol. 17/6, https://doi.org/10.1017/S1355770X12000290.

[70] Henderson, J., A. Storeygard and U. Deichmann (2017), “Has climate change driven urbanization in Africa?”, Journal of Development Economics, Vol. 124, pp. 60-82, https://doi.org/10.1016/j.jdeveco.2016.09.001.

[79] Hertel et al. (2010), “The poverty implications of climate-induced crop yield changes by 2030”, Global Environmental Change, Vol. 20/4, pp. 577-585, https://doi.org/10.1016/j.gloenvcha.2010.07.001.

[153] He, Z. et al. (2017), “Impact of urbanization on energy related CO2 emission at different development levels: Regional difference in China based on panel estimation”, Journal of Cleaner Production, Vol. 140, pp. 1719-1730, https://doi.org/10.1016/j.jclepro.2016.08.155.

[24] Hoegh-Guldberg, O. et al. (2018), Impacts of 1.5°C of Global Warming on Natural and Human Systems, IPCC, https://www.ipcc.ch/sr15/.

[47] Hussein, K. and D. Suttie (2016), Rural-urban linkages and food systems in sub-Saharan Africa, International Fund for Agricultural Development, Rome, http://www.ifad.orgwww.ruralpovertyportal.orgwww.facebook.com/ifadinstagram.com/ifadnewswww.twitter.com/ifadnewswww.youtube.com/user/ifadTV.

[129] Huynh et al (2020), “Climate change vulnerability assessment for Can Tho city by a set of indicators”, International Journal of Climate Change Strategies and Management, Vol. 12/1, pp. 147-158, https://doi.org/10.1108/ijccsm-01-2018-0003.

[143] IEA (2021), Empowering Cities for a Net Zero Future: unlocking resilient, smart, sustainable urban energy systems, International Energy Agency, Paris, https://iea.blob.core.windows.net/assets/4d5c939d-9c37-490b-bb53-2c0d23f2cf3d/G20EmpoweringCitiesforaNetZeroFuture.pdf (accessed on 23 December 2021).

[146] IEA (2018), Global Energy & CO2 Status Report: The latest trends in energy and emissions in 2018, Internatinal Energy Agency, Paris, https://webstore.iea.org/global-energy-co2-status-report-2018.

[144] IEA (2016), Energy Technology Perspectives 2016: Towards Sustainable Urban Energy Systems, International Energy Agency, Paris, https://www.iea.org/reports/energy-technology-perspectives-2016.

[53] IFPRI (2020), 2020 Global Food Policy Report: Building Inclusive Food Systems, International Food Policy Research Institute, Washington DC, https://doi.org/10.2499/9780896293670.

[82] ILO (2019), Working on a warmer planet: The impact of heat stress on labour productivity and decent work, International Labour Office, Geneva, http://www.ilo.org/publns.

[1] IPCC (2021), “Summary for Policy Makers”, in IPCC (ed.), Climate Change 2021: The Physical Science Basis, Cambridge University Press, Cambridge, https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_SPM.pdf (accessed on 25 October 2021).

[164] IPCC (2018), “Summary for Policymakers”, https://www.ipcc.ch/sr15/.

[72] IPCC (2014), Chapter Climate Change 2014 Synthesis Report Summary for Policymakers Summary for Policymakers, IPCC, Geneva, https://www.ipcc.ch/site/assets/uploads/2018/02/AR5_SYR_FINAL_SPM.pdf.

[170] IPCC (2014), Climate change 2014: Mitigation of Climate Change: Working Group III contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, United Kingdom and New York, USA.

[135] Jevrejeva, S. et al. (2018), “Flood damage costs under the sea level rise with warming of 1.5 °C and 2 °C”, Environmental Research Letters, Vol. 13/7, https://doi.org/10.1088/1748-9326/aacc76.

[116] John, L. (2012), Secondary cities in South Africa: The start of a conversation, South African Cities Network, Cape Town.

[44] John, R. (2020), “Flooding in Informal Settlements: Potentials and Limits for Household Adaptation in Dar es Salaam City, Tanzania”, American Journal of Climate Change, Vol. 09/02, https://doi.org/10.4236/ajcc.2020.92006.

[124] Kareem, B. et al. (2020), “Pathways for resilience to climate change in African cities”, Environmental Research Letters, Vol. 15/7, https://doi.org/10.1088/1748-9326/ab7951.

[93] KFW (2015), Climate-Focus-Paper: Cities and Climate Change, https://climate-adapt.eea.europa.eu/metadata/publications/climate-focus-paper-cities-and-climate-change.

[83] Knittel, N. et al. (2020), “A global analysis of heat-related labour productivity losses under climate change—implications for Germany’s foreign trade”, Climatic Change, Vol. 160/2, https://doi.org/10.1007/s10584-020-02661-1.

[77] Knox, J. et al. (2012), “Climate change impacts on crop productivity in Africa and South Asia”, Environmental Research Letters, Vol. 7/3, https://doi.org/10.1088/1748-9326/7/3/034032.

[80] Lecocq, F. and S. Zmarak (2007), “How Might Climate Change Affect Economic Growth in Developing Countries? A Review of the Growth Literature with a Climate Lens”, Policy Research Working Paper, No. 4315, World Bank, Washington DC, http://econ.worldbank.org.

[99] Li, H. et al. (2018), “Interaction between urban heat island and urban pollution island during summer in Berlin”, Science of The Total Environment, Vol. 636, pp. 818-828, https://doi.org/10.1016/j.scitotenv.2018.04.254.

[11] Litman, T. (2014), Analysis of Public Policies That Unintentionally Encourage and Subsidize Urban Sprawl, http://www.lsecities.net.

[98] Li, Y. et al. (2018), “Projecting future climate change impacts on heat-related mortality in large urban areas in China”, Environmental Research, Vol. 163, pp. 171-185, https://doi.org/10.1016/j.envres.2018.01.047.

[58] Lobell et al. (2011), “Climate Trends and Global Crop Production Since 1980”, Science, Vol. 333/6042, pp. 613-616.

[56] Maharjan, K. and N. Joshi (2012), Climate Change, Agriculture and Rural Livelihoods in Developing Countries with Reference to Nepal, Springer Japan, Tokyo, https://doi.org/10.1007/978-4-431-54343-5.

[117] Mahlalela, P. et al. (2020), “Drought in the Eastern Cape region of South Africa and trends in rainfall characteristics”, Climate Dynamics, Vol. 55/9-10, https://doi.org/10.1007/s00382-020-05413-0.

[38] Manda, M. and E. Wanda (2017), “Understanding the nature and scale of risks in Karonga, Malawi”, Environment and Urbanization, Vol. 29/1, https://doi.org/10.1177/0956247817692200.

[28] Marplecroft (2018), 84% of world’s fastest growing cities face “extreme” climate change risks, https://www.maplecroft.com/insights/analysis/84-of-worlds-fastest-growing-cities-face-extreme-climate-change-risks/.

[122] McDonald, R. et al. (2014), “Water on an urban planet: Urbanization and the reach of urban water infrastructure”, Global Environmental Change, Vol. 27, pp. 96-105, https://doi.org/10.1016/j.gloenvcha.2014.04.022.

[23] Mika, J. et al. (2018), Impact of 1.5 K global warming on urban air pollution and heat island with outlook on human health effects, Elsevier B.V., https://doi.org/10.1016/j.cosust.2018.05.013.

[120] Muyambo, F., A. Jordaan and Y. Bahta (2017), “Assessing social vulnerability to drought in South Africa: Policy implication for drought risk reduction”, Jàmbá: Journal of Disaster Risk Studies, Vol. 9/1, https://doi.org/10.4102/jamba.v9i1.326.

[151] Naya, D. (ed.) (2016), “Is Decoupling GDP Growth from Environmental Impact Possible?”, PLOS ONE, Vol. 11/10, p. e0164733, https://doi.org/10.1371/journal.pone.0164733.

[14] NCE (2018), Unlocking The Inclusive Growth Story of the 21st Century: Accelerating Climate Action in Urgent Times, http://www.newclimateeconomy.report.

[96] Ningrum, W. (2017), Urban Heat Island towards Urban Climate, Institute of Physics Publishing, https://doi.org/10.1088/1755-1315/118/1/012048.

[73] NOAA National Centers for Environmental Information (2020), Climate at a Glance: Global Time Series, https://www.ncdc.noaa.gov/cag/.

[54] Nyahunda, L., H. Tirivangasi and H. Tirivangasi (2019), “Challenges faced by rural people in mitigating the effects of climate change in the Mazungunye communal lands, Zimbabwe”, Jàmbá-Journal of Disaster Risk Studies, Vol. 11/1, pp. 1-9, https://doi.org/10.4102/jamba.

[12] OECD (2021), Transport Strategies for Net-Zero Systems by Design, OECD Publishing, Paris, https://doi.org/10.1787/0a20f779-en.

[172] OECD (2021), What Is the Well-Being Lens: An Innovative Process for Net-zero Strategies, OECD Publishing, Paris, https://www.oecd.org/climate-change/well-being-lens/well-being-lens-brochure.pdf (accessed on 20 October 2021).

[169] OECD (2019), Accelerating Climate Action: Refocusing Policies through a Well-being Lens, OECD Publishing, Paris, https://doi.org/10.1787/2f4c8c9a-en.

[167] OECD (2018), Building Resilient Cities: An Assessment of Disaster Risk Management Policies in Southeast Asia, OECD Green Growth Studies, OECD Publishing, Paris, https://doi.org/10.1787/9789264305397-en.

[13] OECD (2018), Rethinking Urban Sprawl: Moving Towards Sustainable Cities, OECD Publishing, Paris, https://doi.org/10.1787/9789264189881-en.

[91] OECD (2016), The Economic Consequences of Outdoor Air Pollution, OECD Publishing, Paris, https://doi.org/10.1787/9789264257474-en.

[166] OECD (2016), Urban Green Growth in Dynamic Asia, OECD Green Growth Studies, OECD Publishing, Paris, https://doi.org/10.1787/9789264266360-en.

[57] OECD (2014), Climate Change, Water and Agriculture: Towards Resilient Systems, OECD Studies on Water, OECD Publishing, Paris, https://doi.org/10.1787/9789264209138-en.

[113] OECD (2012), OECD Environmental Outlook to 2050: The Consequences of Inaction, OECD Publishing, Paris, https://doi.org/10.1787/9789264122246-en.

[27] OECD (2010), Cities and Climate Change, OECD Publishing, Paris, https://doi.org/10.1787/9789264091375-en.

[150] OECD (2002), “Indicators to Measure Decoupling of Environmental Pressure from Economic Growth, OECD Publishing, Paris, https://www.oecd.org/environment/indicators-modelling-outlooks/1933638.pdf (accessed on 11 July 2021).

[147] OECD/IEA (2017), World Energy Outlook 2017, International Energy Agency, Paris, https://www.iea.org/reports/world-energy-outlook-2017.

[39] OECD/UCLG (2016), Subnational Governments around the world: Structure and finance, OECD and UCLG, Paris, https://www.uclg.org/sites/default/files/global_observatory_on_local_finance_0.pdf (accessed on 13 July 2022).

[32] Okaka, F. and B. Odhiambo (2019), “Health vulnerability to flood-induced risks of households in flood-prone informal settlements in the Coastal City of Mombasa, Kenya”, Natural Hazards, Vol. 99/2, pp. 1007-1029, https://doi.org/10.1007/s11069-019-03792-0.

[66] Oliveira, J. and P. Pereda (2020), “The impact of climate change on internal migration in Brazil”, Journal of Environmental Economics and Management, Vol. 103, https://doi.org/10.1016/j.jeem.2020.102340.

[145] Olivier, J. and J. Peters (2020), Trends in Global CO2 and Total Green House Gas Emissions - 2019 Report, PBL Netherlands Environmental Assessment Agency, The Hague, https://www.pbl.nl/sites/default/files/downloads/pbl-2020-trends-in-global-.

[134] Oppenheimer, M. et al. (2019), “Sea Level Rise and Implications for Low-Lying Islands, Coasts and Communities”, in Pörtner, D.C. Roberts, V. Masson-Delmotte, P. Zhai, M. Tignor, E. Poloczanska, K. Mintenbeck, A. Alegría, M. Nicolai, A. Okem, J. Petzold, B. Rama, N. (ed.), IPCC Special Report on the Ocean and Cryosphere in a Changing Climate, https://www.ipcc.ch/srocc/chapter/chapter-4-sea-level-rise-and-implications-for-low-lying-islands-coasts-and-communities/.

[92] Pandey, K. (2010), “Costs of Adapting to Climate Change for Human Health in Developing Countries”, The World Bank, Washington DC, https://openknowledge.worldbank.org/bitstream/handle/10986/27750/581680NWP0Box31CDP0111010Health1web.pdf?sequence=1&isAllowed=y (accessed on 11 July 2021).

[55] Pareek, N. (2017), “Climate Change Impact on Soils: Adaptation and Mitigation”, MOJ Ecology & Environmental Sciences, Vol. 2/3, https://doi.org/10.15406/mojes.2017.02.00026.

[156] Parikh, J. and V. Shukla (1995), “Urbanization, energy use and greenhouse effects in economic development”, Global Environmental Change, Vol. 5/2, pp. 87-103, https://doi.org/10.1016/0959-3780(95)00015-g.

[59] Parker, L. et al. (2019), “Vulnerability of the agricultural sector to climate change: The development of a pan-tropical Climate Risk Vulnerability Assessment to inform sub-national decision making”, PLoS ONE, Vol. 14/3, https://doi.org/10.1371/journal.pone.0213641.

[89] Parncutt, R. (2019), “The human cost of anthropogenic global warming: Semi-quantitative prediction and the 1,000-tonne rule”, Frontiers in Psychology, Vol. 10/OCT, https://doi.org/10.3389/fpsyg.2019.02323.

[35] Paterson, S. et al. (2017), “Size does matter: City scale and the asymmetries of climate change adaptation in three coastal towns”, Geoforum, Vol. 81, pp. 109-119, https://doi.org/10.1016/j.geoforum.2017.02.014.

[15] Pendall, R. (1999), “Do land-use controls cause sprawl?”, Environment and Planning B: Planning and Design, Vol. 26/4, pp. 555-571, https://doi.org/10.1068/b260555.

[119] PMG (2018), Western and Eastern Cape drought crisis, Parliamentary Monitoring Group Water and Sanitation Meeting Summary, https://pmg.org.za/committee-meeting/25770/ (accessed on 12 July 2021).

[65] Podesta, J. (2019), The climate crisis, migration, and refugees, Brookings, Washington DC, https://www.brookings.edu/research/the-climate-crisis-migration-and-refugees/.

[130] Quang Vinh Ky (2018), Planning for peri-urban development and flooding issues: The story of new urban areas in Can Tho City, Institute for Social and Environmental Transition-International, Can Tho, https://www.preventionweb.net/publications/view/58096 (accessed on 12 July 2021).

[131] Rajesh, D. (2018), Managing water amidst urban sprawl and climate change in Can Tho City, Stockholm Environmental Institute, https://www.sei.org/featured/managing-water-amidst-urban-sprawl-climate-change-can-tho-city/ (accessed on 12 July 2021).

[62] Reardon, T. and D. Zilberman (2018), “Climate smart food supply chains in developing countries in an era of rapid dual change in agrifood systems and the climate”, in Natural Resource Management and Policy, Springer, https://doi.org/10.1007/978-3-319-61194-5_15.

[21] Revi et al (2014), Urban Areas in Climate Change 2014: Impacts, Adaptation, and Vulnerability, IPCC, Cambridge, https://www.ipcc.ch/report/ar5/wg2/.

[34] Roberts et al. (2016), Intermediary Cities: The Nexus Between the Local and the Global Co-creating the Urban Future, UCLG, Barcelona, https://www.researchgate.net/publication/309401150.

[140] Roberts, B. (2014), Managing Systems of Secondary Cities, Cities Alliance/UNOPS, Brussels, https://www.researchgate.net/publication/273442050.

[31] Rohat, G. et al. (2019), “Projections of Human Exposure to Dangerous Heat in African Cities Under Multiple Socioeconomic and Climate Scenarios”, Earth’s Future, Vol. 7/5, pp. 528-546, https://doi.org/10.1029/2018EF001020.

[94] Rosenzweig, C. et al. (2015), ARC3.2 Summary for City Leaders. Urban Climate Change Research Network, Columbia University, New York, http://www.uccrn.org/casestudies.

[138] Roy, S. et al. (2018), Bangladesh: National Urban Policies and City Profiles for Dhaka and Khulna, Sustainable, Healthy and Learning Cities and Neighbourhoods (SHLC), http://www.centreforsustainablecities.ac.uk/research/bangladesh-national-urban-policies-city-profiles-dhaka-khulna/ (accessed on 12 July 2021).

[8] Rubiera-Morollón, F. and R. Garrido-Yserte (2020), “Recent Literature about Urban Sprawl: A Renewed Relevance of the Phenomenon from the Perspective of Environmental Sustainability”, Sustainability, Vol. 12/16, p. 6551, https://doi.org/10.3390/su12166551.

[118] SABC (2020), Drought worsening in large parts of Eastern Cape, SABC News, https://www.sabcnews.com/sabcnews/drought-worsening-in-large-parts-of-eastern-cape/ (accessed on 12 July 2021).

[127] Salvadore et al. (2015), “Hydrological modelling of urbanized catchments: A review and future directions”, Journal of Hydrology, Vol. 529, pp. 62-81, https://doi.org/10.1016/j.jhydrol.2015.06.028.

[42] Satterthwaite et al. (2007), Adapting to climate change in urban areas: the possibilities and constraints in low- and middle-income nations, International Institute for Environment and Development, London.

[36] Satterthwaite, D. (2016), “Small and intermediate urban centres in sub Saharan Africa”, Urban Africa Risk Knowledge, No. 6, International Institute for Environment and Development (IIED), London, https://www.urbanark.org/background-paper-small-and-intermediate-urban-centres-sub-saharan-africa (accessed on 11 July 2021).

[148] Satterthwaite, D. (2009), “The implications of population growth and urbanization for climate change”, Environment and Urbanization, Vol. 21/2, pp. 545-567, https://doi.org/10.1177/0956247809344361.

[41] Satterthwaite, D. et al. (2020), “Building Resilience to Climate Change in Informal Settlements”, One Earth, Vol. 2/2, pp. 143-156, https://doi.org/10.1016/j.oneear.2020.02.002.

[16] Savini, F. and M. Aalbers (2016), “The de-contextualisation of land use planning through financialisation: Urban redevelopment in Milan”, European Urban and Regional Studies, Vol. 23/4, pp. 878-894, https://doi.org/10.1177/0969776415585887.

[5] Seto et al. (2012), “Global forecasts of urban expansion to 2030 and direct impacts on biodiversity and carbon pools”, Proceedings of the National Academy of Sciences of the United States of America, Vol. 109/40, pp. 16083-16088, https://doi.org/10.1073/pnas.1211658109.

[163] Seto, K. et al. (2014), “Human Settlements, Infrastructure and Spatial Planning”, in Edenhofer, O. (ed.), Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Mitigation of Climate Change, Cambridge, and New York, https://www.ipcc.ch/site/assets/uploads/2018/02/ipcc_wg3_ar5_chapter12.pdf (accessed on 11 July 2021).

[106] Singh, T., C. Siderius and Y. Van der Velde (2018), “When do Indians feel hot? Internet searches indicate seasonality suppresses adaptation to heat”, Environmental Research Letters, Vol. 13/5, p. 054009, https://doi.org/10.1088/1748-9326/aaba82.

[78] Sultan, B. and M. Gaetani (2016), “Agriculture in West Africa in the Twenty-first Century: climate change and impacts scenarios, and potential for adaptation”, Frontiers in Plant Science, Vol. 7/1262, pp. 1-37.

[75] Swiss Re Institute (2021), The economics of climate change; no action not an option, Swiss Re Management, Zurich, https://www.swissre.com/dam/jcr:e73ee7c3-7f83-4c17-a2b8-8ef23a8d3312/swiss-re-institute-expertise-publication-economics-of-climate-change.pdf (accessed on 23 December 2021).

[71] Tacoli, C. et al. (2015), “Urbanisation, rural-urban migration and urban poverty”, IIED, https://www.researchgate.net/publication/273888747_Urbanisation_Rural-Urban_Migration_and_Urban_Poverty.

[63] Tacoli, C. et al. (2014), World Migration Report 2015. Urbanization, Rural-urban Migration and Urban Poverty, Human Settlements Group International Institute forEnvironment and Development (IIED), London, https://www.iom.int/sites/default/files/our_work/ICP/MPR/WMR-2015-Background-Paper-CTacoli-GMcGranahan-DSatterthwaite.pdf (accessed on 17 February 2021).

[68] Tacoli, C. (2011), Not only climate change: mobility, vulnerability and socio-economic transformations in environmentally fragile areas of Bolivia, Senegal and Tanzania, IIED, London, https://pubs.iied.org/10590IIED/.

[50] Tacoli, C. and B. Vorley (2016), Informal food systems and food security in rural and urban East Africa, https://doi.org/10.1371/journal.

[115] The Straits Times (2019), Water express delivers emergency supplies to drought-hit Indian city, The Straits Times, Asia, https://www.straitstimes.com/asia/south-asia/water-express-delivers-emergency-supplies-to-drought-hit-indian-city (accessed on 25 October 2021).

[30] Tume et al. (2019), “Indigenous Knowledge and Farmer Perceptions of Climate and Ecological Changes in the Bamenda Highlands of Cameroon: Insights from the Bui Plateau”, Climate, Vol. 7/12, p. 138, https://doi.org/10.3390/cli7120138.

[107] Turek-Hankins et al. (2021), “Climate change adaptation to extreme heat: A global systematic review of implemented action”, Oxford Open Climate Change, https://doi.org/10.1093/oxfclm/kgab005.

[49] Turok, I. (2018), “Informing Africa’s urban transformation: A response to Fox et al. and Potts”, Urban Studies, Vol. 55/5, https://doi.org/10.1177/0042098017713575.

[102] Tusholske, C. et al. (2021), Global High Resolution Daily Extreme Urban Heat Exposure (UHE-Daily), 1983-2016, https://doi.org/10.7927/fq7g-ny13.

[26] UCCRN (2018), The Future We Don’t Want -- How Climate Change Could Impact the World’s Greatest Cities, UCCRN, C40, Acclimatise, Global Covenant of Mayors.

[121] UCCRN and C40 (2018), The Future We Don’t Want: How Climate Change Could Impact the World’s Greatest Cities, Columbia University, New York, https://uccrn.ei.columbia.edu/news/future-we-dont-want (accessed on 12 July 2021).

[171] UCLG (2016), UCLG Frame document for Intermediary Cities: Planning and management of sustainable urban development, United Cities and Local Governments , Barcelona, https://www.uclg.org/sites/default/files/uclg_frame_document_ic.pdf (accessed on 18 February 2021).

[142] UN-Climate Action (2018), Cities and Pollution contribute to climate change: Cities and local action, https://www.un.org/en/climatechange/cities-pollution.shtml (accessed on 22 June 2020).

[2] UNDESA (2018), World Urbanization Prospects 2018 Highlights, UNDESA, New York, https://population.un.org/wup/Publications/.

[3] UNDESA (2018), World Urbanization Prospects: The 2018 Revision, Online edition, https://population.un.org/wup/ (accessed on 23 April 2020).

[9] UN-HABITAT (2011), Cities and Climate Change - Global Report on Human Settlements 2011, Un-Habitat, London - Washington DC, https://unhabitat.org/cities-and-climate-change-global-report-on-human-settlements-2011.

[87] UNISDR (2016), The Human Cost of Weather-Related Disasters 1995-2015, UN Office for Disaster Risk Reduction, Geneva, https://www.undrr.org/about-undrr.

[69] Van der Geest, K. (2011), “North-South Migration in Ghana: What Role for the Environment?”, International Migration, Vol. 49/SUPPL.1, https://doi.org/10.1111/j.1468-2435.2010.00645.x.

[61] Vermeulen, S., B. Campbell and J. Ingram (2012), Climate Change and Food Systems, https://doi.org/10.1146/annurev-environ-020411-130608.

[159] Wang et al (2020), “Convergence of carbon intensity: a test on developed and developing countries”, Environmental Science and Pollution Research, Vol. 27/28, https://doi.org/10.1007/s11356-020-09175-4.

[149] Wang, F., F. Yang and L. Qi (2020), “Convergence of carbon intensity: a test on developed and developing countries”, Environmental Science and Pollution Research, Vol. 27/28, pp. 34796-34807, https://doi.org/10.1007/s11356-020-09175-4.

[154] Wang, S. et al. (2018), “The effects of urbanization on CO2 emissions in the Pearl River Delta: A comprehensive assessment and panel data analysis”, Applied Energy, Vol. 228, pp. 1693-1706, https://doi.org/10.1016/j.apenergy.2018.06.155.

[86] Watts, N. et al. (2015), “Health and climate change: policy responses to protect public health”, The Lancet, Vol. 386/10006, https://doi.org/10.1016/S0140-6736(15)60854-6.

[85] Watts, N. et al. (2018), “The Lancet Countdown on health and climate change: from 25 years of inaction to a global transformation for public health”, The Lancet, Vol. 391/10120, https://doi.org/10.1016/S0140-6736(17)32464-9.

[158] Wei, B., H. Yagita and A. Inaba (2003), “Urbanisation impact on energy demand and CO2 emission in China”, Journal of Chonging University, Vol. 2/46, https://www.researchgate.net/publication/284385268_Urbanization_impact_on_energy_demand_and_CO2_emission_in_China/citation/download (accessed on 14 July 2021).

[90] WHO (2018), Climate Change and Health, https://www.who.int/en/news-room/fact-sheets/detail/climate-change-and-health (accessed on 8 July 2020).

[88] WHO (2014), Quantitative risk assessment of the effects of climate change on selected causes of death, 2030s and 2050s, World Health Organisation, Geneva, https://www.who.int/globalchange/publications/quantitative-risk-assessment/en/.

[141] Williams, S. (2021), “The sea is rising, the climate is changing: the lessons learned from Mozambique’s deadly cyclone”, The Guardian, https://www.theguardian.com/world/2021/jan/02/the-sea-is-rising-the-climate-is-changing-the-lessons-learned-from-mozambiques-deadly-cyclone (accessed on 25 October 2021).

[74] WMO (2019), WMO Statement on the State of the Global Climate in 2019, World Meteorological Organization, Geneva, https://library.wmo.int/doc_num.php?explnum_id=10211.

[97] World Bank (2010), Cities and Climate Change: An Urgent Agenda, World Bank, Washington DC, http://documents.worldbank.org/curated/en/194831468325262572/Cities-and-climate-change-an-urgent-agenda.

[60] World Bank, CIAT and CGIAR (2015), Climate-Smart Agriculture in Nicaragua, World Bank, Washington DC.

[139] Zermoglio, F. et al. (2020), Climate vulnerability assessment: Impacts on health outcomes in Secondary Cities of Bangladesh, USAID, New Wark, https://doi.org/10.13140/RG.2.2.22928.15368.

[126] Zhang, W. et al. (2018), “Urbanization exacerbated the rainfall and flooding caused by hurricane Harvey in Houston”, Nature, Vol. 563/7731, https://doi.org/10.1038/s41586-018-0676-z.

[111] Zhang, X. et al. (2019), “Urban drought challenge to 2030 sustainable development goals”, Science of the Total Environment, Vol. 693, https://doi.org/10.1016/j.scitotenv.2019.07.342.

[155] Zhou, Y. et al. (2015), “Effects of rural–urban development transformation on energy consumption and CO2 emissions: A regional analysis in China”, Renewable and Sustainable Energy Reviews, Vol. 52, pp. 863-875, https://doi.org/10.1016/j.rser.2015.07.158.

Annex 2.A. Additional figures and tables

Annex Figure 2.A.1. Expected change in cities’ CO₂ emissions by sector due to changes in GDP, population, and built-up land

Annex Figure 2.A.2. CO₂ emissions per capita across regions (2015)

	(1)	(2)	(3)
	Density decrease	Density decrease	Density decrease

Time to capital	0.978	0.976	0.973
	(-1.58)	(-1.51)	(-1.78)

Border	0.570	0.584	0.793
	(-1.11)	(-0.96)	(-0.41)

River	1.005	1.019	1.051
	(0.03)	(0.11)	(0.30)

Coast, 5Km	0.388^***	0.455^***	0.502^***
	(-4.23)	(-4.29)	(-3.78)

GDP, 2000		1.000	1.000
		(-0.82)	(-0.31)

GDP, 2000-15		0.820^**	0.839^**
		(-2.81)	(-2.89)

< 100K inhab.			5.260^***
			(8.21)

100K-500K inhab.			3.474^***
			(7.47)

500K-1M inhab.			1.920^***
			(4.02)
Country FE	Yes	Yes	Yes
No. Observations	6660	6660	6660
Pseudo R2	0.271	0.302	0.320

Note: Estimates of a logistic econometric model, where the dependent variable indicates whether a city has lost between 2000 and 2015. Covariates include the time to reach the capital city, whether the city is in an international border, whether the city is next to a river basin, whether the city is within 5 Km from the coast, the GDP level in 2000, the GDP growth rate between 2000 and 2015, as well dummy variables indicating whether the city had in 2015 less than 100 000 inhabitants, 100 000 to 500 000 inhabitants, and 500 000 to 1 000 000 inhabitants. All regression models include country fixed-effects. Results based on standard errors clustered by city. Year fixed effects are included. Data for the periods 1990, 2000 and 2015. Exponentiated coefficients; t statistics in parentheses. * p < 0.05, ** p < 0.01, *** p < 0.001.

Source: Author’s own calculations using GHS Urban Centre Database (2019[173]).

	(1)	(2)	(3)
	Log of GDP	Log of GDP	Log of GDP
Log (Number of events over HI 40.6ºC)	-0.649^***	-0.662^***	-0.671^***
	(-9.69)	(-9.86)	(-10.50)

High temperature X Log (Number of events over HI 40.6ºC)	0.588^***	0.604^***	0.618^***
	(7.73)	(7.94)	(8.40)

50K-100K X Log (Number of events over HI 40.6ºC)	0.284^***	0.287^***	0.282^***
	(3.96)	(3.97)	(4.08)

100K-500K X Log (Number of events over HI 40.6ºC)	0.202^**	0.202^**	0.203^**
	(2.83)	(2.81)	(2.95)

500K-1M X Log (Number of events over HI 40.6ºC)	0.189	0.175	0.176
	(1.92)	(1.76)	(1.83)

High temperature X 50K-100K X Log (Number of events over HI 40.6ºC)	-0.325^***	-0.335^***	-0.331^***
	(-4.00)	(-4.10)	(-4.18)

High temperature X 500K-1M X Log (Number of events over HI 40.6ºC)	-0.197^*	-0.203^*	-0.204^**
	(-2.45)	(-2.51)	(-2.60)

High temperature X 500K-1M X Log (Number of events over HI 40.6ºC)	-0.177	-0.165	-0.167
	(-1.62)	(-1.50)	(-1.55)

Log of Built up		0.105^***	0.104^***
		(5.73)	(5.69)

Log of Population		0.146^***	0.145^***
		(4.06)	(4.02)

Log of average temperature (5 years)			4.622^***
			(6.12)

Constant	18.75^***	17.09^***	0.628
	(1204.99)	(42.39)	(0.23)
City FE	Yes	Yes	Yes
Year FE	Yes	Yes	Yes
No. Observations	20830	20765	20744
N of Groups	7625	7615	7606
R-sq Within	0.707	0.716	0.717
R-sq Between	0.00455	0.287	0.00559
R-sq Overall	0.0505	0.308	0.0531
	(1)	(2)	(3)
Main effects	Log of GDP	Log of GDP	Log of GDP

Low temperature X 50K-100K	-0.365^***	-0.375^***	-0.390^***
	(-13.89)	(-13.80)	(-14.52)

Low temperature X 100K-500K	-0.447^***	-0.459^***	-0.469^***
	(-17.77)	(-17.36)	(-18.10)

Low temperature X 500K-1M	-0.460^***	-0.486^***	-0.496^***
	(-6.31)	(-6.60)	(-6.89)

Low temperature X >1M	-0.649^***	-0.662^***	-0.671^***
	(-9.69)	(-9.86)	(-10.50)

High temperature X 50K-100K	-0.102^***	-0.106^***	-0.102^***
	(-6.81)	(-7.35)	(-7.05)

High temperature X 100K-500K	-0.0565^***	-0.0587^***	-0.0551^***
	(-4.19)	(-4.46)	(-4.15)

High temperature X 500K-1M	-0.0492	-0.0475	-0.0442
	(-1.51)	(-1.49)	(-1.37)

High temperature X >1M	-0.0611	-0.0577	-0.0533
	(-1.73)	(-1.64)	(-1.50)
Observations	20830	20765	20744

Note: Estimates of a panel econometric model regressing the Log of GDP on: Log of number of days that exceeded a HI>40.6ºC interacted by city size and whether the city is high or low temperature. Other variables included as controls are log of population, log of built-up area and log of average temperature (calculated over the last 5 years). Results based on a fixed-effects estimator and standard errors clustered by city. Year fixed effects are included. Data for the periods 1990, 2000 and 2015. t statistics in parentheses. * p < 0.05, ** p < 0.01, *** p < 0.001.

Source: Author’s own calculations using GHS Urban Centre Database (2019[173]) and (Tusholske et al., 2021[102]).

	(1)	(2)	(3)
	Log of GDP	Log of GDP	Log of GDP
Log (Number of events over WB 30ºC)	-0.559^***	-0.562^***	-0.585^***
	(-6.35)	(-6.39)	(-6.79)

High temperature X Log (Number of events over WB 30ºC)	0.519^***	0.527^***	0.554^***
	(5.54)	(5.61)	(5.99)

50K-100K X Log (Number of events over WB 30ºC)	0.258^**	0.254^**	0.244^**
	(2.82)	(2.77)	(2.72)

100K-500K X Log (Number of events over WB 30ºC)	0.113	0.0991	0.102
	(1.23)	(1.08)	(1.14)

500K-1M X Log (Number of events over WB 30ºC)	0.162	0.134	0.132
	(1.28)	(1.07)	(1.08)

High temperature X 50K-100K X Log (Number of events over WB 30ºC)	-0.302^**	-0.305^**	-0.293^**
	(-3.09)	(-3.10)	(-3.04)

High temperature X 500K-1M X Log (Number of events over WB 30ºC)	-0.115	-0.106	-0.107
	(-1.19)	(-1.08)	(-1.11)

High temperature X 500K-1M X Log (Number of events over WB 30ºC)	-0.182	-0.150	-0.145
	(-1.37)	(-1.14)	(-1.13)

Log of Built up		0.0996^***	0.0963^***
		(5.25)	(5.07)

Log of Population		0.157^***	0.158^***
		(4.20)	(4.24)

Log of average temperature (5 years)			6.885^***
			(8.19)

Constant	18.73^***	16.94^***	-7.609^*
	(1092.08)	(40.38)	(-2.53)
City FE	Yes	Yes	Yes
Year FE	Yes	Yes	Yes
No. Observations	19902	19842	19822
N of Groups	7455	7446	7437
R-sq Within	0.707	0.716	0.718
R-sq Between	0.00494	0.265	0.000221
R-sq Overall	0.0454	0.297	0.0118
	(1)	(2)	(3)
Main effects	Log of GDP	Log of GDP	Log of GDP

Low temperature X 50K-100K	-0.301^***	-0.308^***	-0.341^***
	(-11.95)	(-11.80)	(-12.88)

Low temperature X 100K-500K	-0.446^***	-0.463^***	-0.483^***
	(-18.24)	(-17.83)	(-18.87)

Low temperature X 500K-1M	-0.397^***	-0.429^***	-0.454^***
	(-4.33)	(-4.88)	(-5.27)

Low temperature X >1M	-0.559^***	-0.562^***	-0.585^***
	(-6.35)	(-6.39)	(-6.79)

High temperature X 50K-100K	-0.0833^***	-0.0857^***	-0.0805^***
	(-5.69)	(-6.15)	(-5.75)

High temperature X 100K-500K	-0.0424^**	-0.0420^**	-0.0357^**
	(-3.12)	(-3.15)	(-2.68)

High temperature X 500K-1M	-0.0595^*	-0.0511	-0.0448
	(-2.19)	(-1.83)	(-1.59)

High temperature X >1M	-0.0395	-0.0351	-0.0312
	(-1.25)	(-1.09)	(-0.95)
Observations	19902	19842	19822

Note: Estimates of a panel econometric model regressing the Log of GDP on: Log of number of days that exceeded a WB>30ºC interacted by city size and whether the city is high or low temperature. Other variables included as controls are log of population, log of built up area and log of average temperature (calculated over the last 5 years). Results based on a fixed-effects estimator and standard errors clustered by city. Year fixed effects are included. Data for the periods 1990, 2000 and 2015. t statistics in parentheses. * p < 0.05, ** p < 0.01, *** p < 0.001.

Source: Author’s own calculations using GHS Urban Centre Database (2019[173]) and (Tusholske et al., 2021[102]).

	Log of CO₂ total	Log of CO₂ residential	Log of CO₂ transport	Log of CO₂ industrial
Log of GDP	0.377^***	0.0825^***	0.489^***	0.313^***
	(15.09)	(3.68)	(20.25)	(13.83)

50K-100K* Log of GDP	-0.249^***	-0.116^***	-0.148^***	-0.0857^***
	(-9.69)	(-4.71)	(-5.77)	(-3.77)

100K-500K* Log of GDP	-0.183^***	-0.122^***	-0.0761^**	-0.0572^*
	(-7.10)	(-5.02)	(-2.92)	(-2.50)

500K-1M* Log of GDP	0.0161	-0.0669^*	0.000862	-0.0243
	(0.44)	(-2.56)	(0.03)	(-0.82)

Log of Population	0.190^**	0.331^**	-0.250^***	0.412^***
	(3.26)	(2.77)	(-4.25)	(5.75)

50K-100K*Log of Population	-0.0401	-0.0937	0.0378	0.0348
	(-0.65)	(-0.76)	(0.61)	(0.41)

100K-500K*Log of Population	0.0258	0.0476	0.00525	0.0360
	(0.42)	(0.39)	(0.08)	(0.48)

500K-1M*Log of Population	-0.0855	0.125	-0.0357	-0.0740
	(-1.10)	(0.96)	(-0.39)	(-0.67)

Log of Built up	0.0190	-0.00195	-0.0166	-0.0508
	(0.41)	(-0.03)	(-0.32)	(-0.91)

50K-100K*Log of Built up	0.161^***	0.346^***	0.111^*	0.0971
	(3.33)	(4.88)	(2.14)	(1.69)

100K-500K*Log of Built up	0.131^**	0.332^***	0.0762	0.0959
	(2.67)	(4.62)	(1.46)	(1.71)

500K-1M*Log of Built up	-0.0103	0.0328	0.0321	0.0835
	(-0.17)	(0.43)	(0.44)	(1.17)
City FE	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
No. Observations	33608	33608	32844	33608
N of Groups	11393	11393	11138	11393
R-sq Within	0.515	0.225	0.760	0.623
R-sq Between	0.461	0.446	0.459	0.545
R-sq Overall	0.460	0.433	0.459	0.535
	(1)	(2)	(3)	(4)
Main effects	Log of CO₂ total	Log of CO₂ residential	Log of CO₂ transport	Log of CO₂ industrial
Log of GDP
50K-100K	0.128^***	-0.0333^*	0.341^***	0.227^***
	(12.65)	(-2.44)	(25.41)	(23.27)

100K-500K	0.194^***	-0.0392^**	0.413^***	0.256^***
	(19.71)	(-3.22)	(28.69)	(22.92)

500K-1M	0.394^***	0.0156	0.490^***	0.289^***
	(14.35)	(1.00)	(18.86)	(13.42)

>1M	0.377^***	0.0825^***	0.489^***	0.313^***
	(15.09)	(3.68)	(20.25)	(13.83)
Log of Population
50K-100K	0.150^***	0.238^***	-0.212^***	0.446^***
	(6.65)	(7.16)	(-10.83)	(8.32)

100K-500K	0.216^***	0.379^***	-0.245^***	0.448^***
	(9.01)	(10.88)	(-11.98)	(16.54)

500K-1M	0.105^*	0.456^***	-0.286^***	0.338^***
	(2.03)	(8.27)	(-4.06)	(3.92)

>1M	0.190^**	0.331^**	-0.250^***	0.412^***
	(3.26)	(2.77)	(-4.25)	(5.75)
Log of Built up
50K-100K	0.180^***	0.344^***	0.0945^***	0.0463^**
	(11.40)	(12.65)	(8.01)	(3.00)

100K-500K	0.150^***	0.330^***	0.0596^***	0.0451^***
	(7.54)	(10.45)	(4.34)	(3.44)

500K-1M	0.00871	0.0308	0.0155	0.0327
	(0.23)	(0.81)	(0.29)	(0.69)

>1M	0.0190	-0.00195	-0.0166	-0.0508
	(0.41)	(-0.03)	(-0.32)	(-0.91)
Observations	33608	33608	32844	33608

Note: Estimates of a panel econometric model regressing CO₂ emissions (total, residential, transport or industrial) on: GDP, population, and built-up, the three variables interacted with city size. Base category is cities above 1 million inhabitants. All variables included expressed logarithmic terms. Results based on a fixed-effects estimator, considering year-effects, and standard errors clustered by city. Data for the periods 1990, 2000 and 2015. t statistics in parentheses. * p < 0.05, ** p < 0.01, *** p < 0.001.

Source: Authors’ own calculations using GHS Urban Centre Database (2019[173]).

Notes

← 1. The Working Group I contribution to the Sixth Assessment Report published in 2021.

← 2. The ratios were obtained through a Google Scholar search. For the overall comparison, the searches were: (i) "secondary city" OR "secondary cities" OR "intermediary city" OR "intermediary cities" OR "small city" OR "small cities" OR "medium cities" OR "medium city" vs (ii) "capital city" OR "capital cities" OR "metropolis" OR "megacity" OR "megacities" OR "big city" OR "big cities". For the climate change topic comparison; (iii) "climate change" AND ("secondary city" OR "secondary cities" OR "intermediary city" OR "intermediary cities" OR "small city" OR "small cities" OR "medium cities" OR "medium city") vs (iv) "climate change" AND ("capital city" OR "capital cities" OR "metropolis" OR "megacity" OR "megacities" OR "big city" OR "big cities").

← 3. To answer this question, we analyse recent estimates from the GHS Urban Centres Database (GHS-UCD) covering different time periods: 1990, 2020 and 2015. This dataset addresses some important caveats from other existing datasets (including the ones from the UNDESA presented above). First, this dataset is not based on administrative boundaries; instead, it uses a standardised definition for cities that improves international comparison. Second, it provides estimates of cities with at least 50 000 inhabitants (estimates from the UN do not account for cities with fewer than 300 000 inhabitants). This is particularly important for less developed regions, where large shares of the population reside in cities with fewer than 300 000 people.

← 4. The 7% in annual GDP is calculated by adding up the annual external cost (USD 400 billion, cost of loss in fitness and public health), and the annual internal cost (USD 625 billion, costs borne by commuters), and as % of total US GDP in 2015 (USD 18 trillion).

← 5. Based on projection of urban population to reach 3.5 billion across the 970 cities in the study, by 2050.

← 6. In 2001 and 2019.

← 7. Based on baseline period 1961-90.

← 8. Cities are classified as warm or cold depending on whether their average temperature was above or below the median temperature in the sample in 1990.

← 9. Chennai is a far larger city than what is considered an intermediary city in this study.

← 10. Under the very low GHG emissions pathway leading to a rise in global temperatures of 1.4°C by 2081-2100.

← 11. Urban areas that are located within 5 km of the coast and that have an elevation of less than 50 m are highly exposed to sea-level rise and coastal flooding.

← 12. The sectors include: energy (power industry), residential (energy for buildings, solid and water waste), industry (combustion for manufacturing, oil refineries and transformation industry, fuel exploitation, industrial processes, solvents and products use), transport (road, off-road, shipping and aviation) and agriculture.

← 13. CO2 emissions are modelled as a function of income, population, and built-up land, i.e. CO2 = f(GDP, Population, Built-up). This specification for the dependent variable was preferred over CO2 per capita in order to identify the effect of population on emissions.

← 14. To test whether the effects of the four city sizes are significantly different, Wald tests have been performed.

← 15. Defined as the percentage of the population living in urban areas.

← 16. According to the same report, 72% of the mitigation potential identified depends directly or indirectly on collaboration with higher levels of government.

← 17. An industrial city of 2 million inhabitants.

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment

Governance

Health

Industry, business and entrepreneurship

Regional, rural and urban development

Science, technology and innovation

Society

Taxation

Trade

Energy

Nuclear energy

Transport

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment

Governance

Health

Industry, business and entrepreneurship

Regional, rural and urban development

Science, technology and innovation

Society

Taxation

Trade

Energy

Nuclear energy

Transport

Countries A - C

Countries D - I

Countries J - M

Countries N - R

Countries S - T

Countries U - Z

Regional and global engagement

Countries

Countries A - C

Countries D - I

Countries J - M

Countries N - R

Countries S - T

Countries U - Z

Regional and global engagement

Publications

Publications

Featured publications

Data

Data

Featured data

News & Events

News & Events

Featured Events

About OECD

About

Engage with us

Work with us

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment