Economic risk assessment of climate change at the city level. The case of Cape Town, South Africa

ABSTRACT

Estimating the economic risks of climate shocks and climate stressors on spatially heterogeneous cities over time remains highly challenging. The purpose of this paper is to present a practical methodology to assess the economic risks of climate change in middle-income cities to inform municipal climate response strategies. Building on a capital-based framework (CBF), spatially disaggregated baseline and future scenario scores for economic wealth and its exposure to climate change are developed for six different classes of capital across 77 major suburbs in Cape Town, South Africa. Capital-at-risk was calculated by combining relative exposure and capital scores across different scenarios, with population impacted for major suburbs and the city’s eight main planning districts. The economic risk assessment presented here provides a generic approach to assist city managers through an enhanced understanding of the relative levels of capital endowment across the city vis-à-vis relative levels of exposure to climate-related hazards over time.

KEYWORDS:

• Climate change
• economic risk assessment
• capital-based framework
• six-capital framework
• Cape town

Introduction

Risk assessment of climate variability, climate change and natural disasters tend to focus on the hazard itself, the exposure to hazards, and vulnerability to inform disaster risk management (Walter 2012). Climate change risk assessment focusses on the consequences and likelihood of, and responses to, the impacts of climate change (Adger et al. 2018) and is conceptualised as a ‘fundamental interaction of the physical risk with the societal process of prioritising, avoiding harm and making legitimate decisions’ (Adger et al. 2018, p. 19). Economic assessments in urban contexts generally compare costs and benefits for the impacts of and responses to climate change and have focussed on specific risks such as flooding (Zhou et al. 2012; Tang et al. 2021; Ding et al. 2022) and fire (Michael et al. 2018), on infrastructure development (Sharma et al. 2009), on climate adaptation options (Zhou et al. 2012; Cartwright et al. 2013) and on city policies for responding to climate change impacts (Estrada et al. 2017).

Hazard and vulnerability assessments as well as economic cost–benefit analyses on impacts and responses are useful (IPCC 2022) but do not answer the question of how much an urban economy is at risk to climate change. Of particular concern in this paper is how socio-economic development at the city level is at risk to climate change, and how this would inform municipal climate response strategies that result in inclusive, equitable, resilient, and spatially integrated cities through a just urban transition process (SACN 2022). Several models and approaches to inform economic risk assessment on an urban level are available in the literature, but these either lack spatial resolution or are not yet well-developed, especially in middle-income cities. For example, scenario development is one approach that helps to construct narratives on impacts and costs for alternative futures under climate change, but generally lacks the detailed quantification of costs and spatial resolution needed for more directed responses (Cartwright et al. 2013). Another example is economic risk assessments that would theoretically require a probabilistic and spatially sensitive integrative modelling approach on an urban scale (Lindley et al. 2007) – an approach especially relevant for spatially heterogeneous, highly unequal and middle-income cities like Cape Town. Accounting for spatially sensitive interlinkages and indirect damages and losses in economic assessments on an urban level would, however, require spatially sensitive macroeconomic models, which in turn are reliant on updated input – output tables or social accounting matrices – all of which are not always readily available at city level (Ye et al. 2021). Published quantitative economic assessments of climate risk generally tend to focus on global, national, or regional levels, with only a few studies attempting to examine climate risk at the city level (Ye et al. 2021; IPCC 2022) – and yet cities are responsible for most of the materials consumed and thus stand to stymie the objectives of sustainable development (SACN 2022). For example, and relevant to our case, economic modelling work has been done on assessing the economic impacts of climate change for the Western Cape Province, but no actual down-scaled climate modelling results were available, and the results are focussed on the economic impacts only on provincial economic sectors (Cartwright et al. 2018). In general, despite some progress, ‘… accurate quantitative assessment of city-specific climate change risk remains highly challenging’ (Ye et al. 2021). The absence of sufficient data and reliable quantitative economic assessment models at the city level necessitates alternative practical approaches to support decision-making on climate change. This paper aims to develop a practical approach for spatially sensitive decision-making on the economic risks of climate change at the city level.

City managers are faced with the question of how to allocate scarce resources over space and time, across multiple competing priorities, exacerbated by climate risks. Decisions on the allocation of resources have to be made in contexts of spatial heterogeneity in socio-economic vulnerability as well as climate risks that are expected to vary significantly over time. Although theoretically a possibility, practically usable economic risk assessment tools and models on an urban level are not readily available. From the perspective of a city manager, there is thus a need for a robust, practical and spatially sensitive screening tool that highlights the economic capital-at-risk under various climate scenarios. The results of such a screening would in turn be used to inform climate response strategies. The approach reported on in this paper was to link climate model results on climate exposure at the city level with various categories of capital indicators as approximations for a city’s economic wealth. One of the challenges is to develop a realistic and balanced selection of indicators, which does bring issues of selection and its weighing, but, as must be stated from the onset, for which there is ‘no scientifically valid solution’ (Dahl 2007).

An earlier Hazard, Vulnerability and Risk Assessment (HVRA) conducted for Cape Town identified key climate driven risks (stressors and shocks) by modelling climate change indicators over time (Petrie et al. 2019). The assessment created a vulnerability index to determine different levels of risk and resilience across spatially discrete locations throughout Cape Town. Specifically, the HVRA anticipates a drier and warmer climate for the City of Cape Town, while the top climate-related hazards facing the city include drought, fire, heatwaves, floods, and strong winds. The water supply system in the Western Cape Province, which supplies the City of Cape Town with the majority of its bulk water requirements, faces significant risk from repeated drought events, with indirect impacts of a water availability crisis affecting the economy, environment, and people of the City of Cape Town. The findings of the HVRA also corroborate findings from other research studies. An earlier study on changes in rainfall and precipitation likewise included forecasts of a decrease in wet days, an increase in dry spells, and a decrease in annual precipitation (Abiodun et al. 2017). Combined with increased temperatures and evaporation, decreased water availability in the medium- to long-term is probable and changes to the City of Cape Town’s water resource management regime in the light of climate change are necessary and ongoing (City of Cape Town 2020a). Moreover, fires pose a direct threat to human life and assets. Informal settlements are also more vulnerable to increased flooding and the risk of fires given their location and limited access to services and resources (Mukheibir and Ziervogel 2006). The climate risk to the City of Cape Town as a coastal city further includes a rise in sea level, which threatens infrastructure as well as the real estate and tourism industries (Colenbrander et al. 2015). With much of the City of Cape Town’s industrial, commercial, and residential areas lying below 10 m above sea level, a rise in sea level increases the vulnerability of beaches and coastal developments in the city. Improved stormwater strategies may be necessary as a rise in sea level, storm surge and heavy rainfall are expected to exacerbate water pollution, compromise drinking water and damage coastal treatment plants (Kessler 2011). Climate shocks and stressors are posing physical risks to the Western Cape (including the City of Cape Town) that will pose risks to the economy as well (Midgley et al. 2005; Cartwright et al. 2012; City of Cape Town 2022).

While the nature and real physical impact of climate shocks and stressors to the City of Cape Town have now been determined using the HVRA, the need for estimations on spatially explicit economic risks of climate change to the city is becoming more pertinent. The theoretical framework used in this paper is a ‘capital-based framework’ (CBF) (Wu and Wu 2012) in support of an assessment of climate risks to the urban economy as a whole. Our study applied a six-capital framework (financial, manufactured, human, intellectual, social, and natural) to provide a holistic view of economic wealth at risk due to climate events and stressors. The six capitals theoretical framework is in development in the fields of governance and integrated reporting (Adams et al. 2013; Dumay et al. 2017) in the wake of increasing demand for non-financial reporting after the financial crisis of 2008/9. The six capitals theoretical framework has been applied in various forms in fields as diverse as corporate social responsibility (CSR) (Fordham et al. 2018) and rural development (Mikulcak et al. 2015). In the field of climate adaptation, multiple-capital framework approaches have been applied to measuring coping and adaptive capacity of cities (Tinch et al. 2015). The relevance of integrated reporting will increase as specific directives for non-financial reporting on social and environmental impacts are transformed into notional law, as has happened in the European Union already (Velte and Stawinoga 2017).

It must be noted that even though the CBF overlaps with, it also differs in many respects from the capital theory approach (CTA) as advanced in economic theory (Hartwick 1977; Solow 1986). The CTA has been developed into indicators for sustainability in its weak (Pearce and Atkinson 1993; Hamilton and Clemens 1999) and strong (Costanza and Daly 1992; Ekins et al. 2003) forms and has been promoted as a tool for sustainability policy (Lange et al. 2018; Tzachor 2020). The urban economy is portrayed in terms of the different types of capital (or assets) that provide a flow of benefits or services to the economy and society; that is, they are defined as ‘productive assets available to the economy’ (Stern 1997). Sustainability is then defined as maintaining a constant aggregate capital stock over time (Hartwick 1977; De Wit and Blignaut 2000). The decision rule to achieve sustainability would be that once capital is used to produce or consume, the rents (over and above an acceptable level of profit) should be reinvested to ensure that the same level of capital will be available for posterity. Questions on the relative importance of various categories of capitals in economic growth and development, and the substitutability between these capitals has dominated much of the literature on the CTA to sustainability and its critics (Stern 1997; De Wit and Blignaut 2000). In this paper, we start with the same premise as the CTA, namely that various productive assets are available to the economy but give a more inclusive and comprehensive representation of all capital asset categories (also non-financial) in advancing the ability of city managers to respond to climate risks in urban economies. As market prices often do not include the damages caused by excessive productivity in one capital category on others (e.g. environmental damage or decline in social fabric as a result of excessive financial accumulation or exploitative labour practices), these cannot be treated as a satisfactory benchmark for calculating the aggregate value for each type of capital. Given the absence of monetary values for all capitals present in a city’s economy, it follows that the relative importance of each of the capitals cannot singularly be derived from any particular model of the economy that only includes monetised values. In our approach, we have therefore not relied on monetised indicators only. Furthermore, as we are not following an economic modelling approach in this paper, the otherwise very important question of substitutability between various capitals over time is excluded from our analysis. Therefore, although there are important overlaps with the CTA, most notably that the stocks of a capital need to be maintained over time to progress towards sustainability, our approach cannot theoretically be categorised as such.

In summary, the purpose of this paper is to develop a practical and spatially sensitive economic risk assessment approach to climate change at an urban scale. This will be achieved by developing a set of spatially sensitive indicators on six capitals at risk at the city level, coupled with scenarios on how they are likely to be affected across various levels of climate exposure over time. We expect that the development of such an approach will enhance decision-making on climate risk at the city level, especially in cities characterised by large spatially sensitive socio-economic inequalities, such as the City of Cape Town.

Methodology

Case study: the City of Cape Town

The City of Cape Town, located on the south-western tip of South Africa, is the country’s second most populous city and the economic hub of the Western Cape Province (see Figure 1 for a map). The City of Cape Town is well-known for its harbour and natural environment (situated within the world-renowned Cape Floristic Region), and for landmarks such as Table Mountain and Cape Point. However, like most of South Africa, the socio-economic dynamics of the City of Cape Town present a story of stark inequality. According to (Statistics South Africa [date unknown]), almost 36% of the population of the city live below of the poverty line, with a dependency ratio in excess of 43%. These inequalities manifest clearly in the spatial distribution of the population, a legacy of the Apartheid-era spatial planning policies. The City of Cape Town also faces a variety of climate change challenges including significant increases in temperature, long-term decreases in rainfall, changes in rainfall seasonality, more extreme heat days and heatwaves, and coastal erosion (Petrie et al. 2019). The combination of these climate change threats and the prevailing socio-economic dynamics presents complex risk management scenarios to city managers.

Figure 1. Map of the City of Cape Town.

Source: Esri, CGIAR | Esri South Africa, Esri, HERE, Garmin, FAO, METI/NASA, USGS

Scenarios

Climate change-related hazards occur at two distinct time scales, namely as events and disasters in the short-term (climate shocks) and as gradual changes over longer time periods (climate stressors). Future risk pathways between climate change and the economy are conceptualised to include both climate exposure (including climate shocks and stressors) and projections of economic wealth at risk as measured on a scale of weakest to strongest indicators for each of the six capitals and their equally weighted aggregates. The climate change projections used were derived from the Climate Change Hazard, Vulnerability, and Risk Assessment Report for the City of Cape Town (Petrie et al. 2019). Each aggregate score was weighted equally based on the outcomes of focus groups and workshops with the City of Cape Town staff in the absence of any scientific evidence base to justify weighting these scores differently (OneWorld 2018). The outcome is that four spatially explicit scenarios are used to illustrate alternative futures for (i) high- and low-bound climate exposure and (ii) weak and strong capital projections.

Indicators and data

The following six forms of capital are included in our economic risk assessment:

• Financial capital: The pool of funds available including debt, equity, and grants generated through private and public operations and investments.

• Manufactured (durable) capital: Includes tools, machinery, buildings, equipment, and other infrastructure (roads, bridges, ports, railways, and water and treatment plants).

• Human capital: Investment in skills, education and training determining the individual’s competencies, capabilities, training, and overall productivity.

• Intellectual capital: Intellectual property (patents, software, copyrights, and licences), organisational capital (tacit knowledge, protocols, and systems), and other intangibles (city brand and reputation).

• Social capital: Institutions and customs organising economic activity (shared norms, common values, non-physical culture, trust, and willingness to engage).

• Natural capital: Natural systems including atmosphere, lithosphere, hydrosphere, and biosphere.

Typical indicators and metrics for each of the six types of capital were developed and are included in Appendix A. As not all indicators were equally well-developed spatially in the City of Cape Town’s databases, a representative set of indicators were adopted with participants from the City of Cape Town’s management. The question of which leading or main indicator(s) and/or composite indicators to include as a measure of the various capitals did not reside with the research team only and was verified with participants from the City of Cape Town management. Leading indicators refer to the use of individual metrics to broadly represent a stock and/or flow of a particular system (i.e. the number of employed people can be considered a leading indicator for human capital). A composite indicator is a function of two or more metrics combined to represent a specific aspect of a system (i.e. combining metrics of biodiversity and vegetation cover to represent natural capital). The indicator approach was validated with participants from the City of Cape Town management, which necessitated the use of both leading and composite indicators to represent the six types of economic capital. Metrics for these indicators were used to represent the baseline scenario, which provided the basis for projecting future scenarios. The main indicators that were ultimately selected and how they are measured for each capital category are included in Table 1.

Table 1. Metrics and projections for six capital indicators.

Capital	Main indicator(s)	Measure	Projections (strong/weak)
Financial	Total revenue	Total budget expenditure (rands)	↑ 2.4%; ↓ 2.4%
Human	Employment	# of employed people	↑ 5%; ↑ 2%
Natural	Ecosystem functioning	Ecosystem services index*	↑ 3%; ↓ 3%
Social	Crime rate	# of crimes	↓ 2%; ↑ 2%
Manufactured	Residential property value	Mean property value (rands)	↑ 10%; ↑ 5%
	Commercial and industrial property value	Mean property value (rands)	↑ 10%; ↑ 5%
	Access to critical infrastructure	Mean travel time to hospitals and fire stations (minutes)	↑ 2%; ↓ 2%
Intellectual	Education levels	# of people with higher education	↑ 3%; ↑ 1%

*An index value was used to represent natural capital rather than an individual lead indicator because of the variety of ecosystem types that function within the urban extent of the City of Cape Town. This index represents a combination of ecosystem intactness and integrity indicators for biodiversity, forestry and watercourses and wetlands..

Economic risk assessment

Risk assessment traditionally involves an estimation of the magnitude of potential consequences or the levels of impacts, and the likelihood or the levels of probability of such impacts happening. The spatial and temporal extent of the risks can also be evaluated in such an assessment. We assessed economic risks based on the outcomes of results from the spatially explicit HVRA (Petrie et al. 2019). For each climate shock and stressor (or composite of such stressors and events, i.e. exposure) impacting on the capitals used in the economy, the economic risks needed to be assessed. The following stepwise approach was followed:

Step 1: Develop a ‘baseline’ for the six capitals (2018 data) that are already functional in the City of Cape Town. The output in the productive economy is dependent on the well-functioning of these six types of capital as measured by selected indicators as indicated in Table 1.

Step 2: Develop alternative futures for the capitals on a continuum of relatively weak capital functioning to strong capital functioning over time. Two future projections were generated for each capital indicator to represent a realistic ‘strong’ and ‘weak’ capital future. The last column of Table 1 indicates the strong and weak projections for the six capital categories. The projections were based on historical trends for these variables and the extent of realistic worst-case and best-case projections (OneWorld 2018). The data used to represent each capital metric was projected until 2030 and normalised against a range from 0 to 1, to ensure commensurability of the different metrics. The min-max feature scaling produces a range of values between 0 and 1. The minimum value is subtracted (from the indicator value) and divided by the difference between the maximum and minimum values. The reason for the development of a common scale when developing a composite indicator is that different socio-economic indicators do not have the same unit of measurement and differ in scale. An overall capital score was derived by equally weighing the six capital scores. To ensure spatial commensurability and to ensure that the spatial dimensions align closely with the approach followed in the HVRA, all capital scores were computed at the major suburb level in the City of Cape Town. A map of the major suburbs in the City of Cape Town is included in Appendix B. (Normalised capital scores per major suburb are included in the supplementary material to this paper).

Step 3: Overlay projections on alternative climate change exposure futures with projections on alternative capital futures to identify areas where the economy, as measured through the indicators for capitals, is at relatively higher and lower risk in the short- to medium-term future. The overall capital score was combined with the medium-term future exposure index variables (derived from Petrie et al. (2019)) to indicate the areas where capital is most at risk within the City of Cape Town. Table 2 outlines the different exposure variables that determine the exposure index, as well as approximated high and low bounds of uncertainty associated with medium-term future projections on these exposures. Uncertainty bounds were used to develop measures of high and low exposure for the medium-term future.

Table 2. Exposure index variables and medium-term exposure bounds.

Exposure variables	Medium-term exposure
	High Bound	Low Bound
Average, maximum and minimum temperature	↑ 3°C	↑ 1°C
Very hot days	↑ 20 days	↑ 0 days
Heat-wave days	↑ 10 days	↑ 0 days
High fire-danger days	↑ 20 days	↑ 0 days
Rainfall	↓ 120 mm	↓ 60 mm
Extreme rainfall	↓ 3 days	↓ 0 days

Source: Based on Petrie et al. (2019).

Step 4: A combined capital and exposure assessment was undertaken for four different spatially explicit scenarios, namely:

• Scenario 1: high bound climate exposure; weak capitals

• Scenario 2: high bound climate exposure; strong capitals

• Scenario 3: low bound climate exposure; weak capitals

• Scenario 4: low bound climate exposure; strong capitals

Capital-at-risk

Capital-at-risk for each of the major suburbs was derived by combining relative exposure and capital scores for all scenarios as per the following equation:

(1) $\mathrm{C}\mathrm{a}\mathrm{p}\mathrm{i}\mathrm{t}\mathrm{a}\mathrm{l}-\mathrm{a}\mathrm{t}-\mathrm{r}\mathrm{i}\mathrm{s}\mathrm{k}=\mathrm{E}\mathrm{x}\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{u}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{c}\mathrm{o}\mathrm{r}\mathrm{e}\times \mathrm{C}\mathrm{a}\mathrm{p}\mathrm{i}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{s}\mathrm{c}\mathrm{o}\mathrm{r}\mathrm{e}$(1)

Results

Six capitals assessment

Figure 2 shows the maps illustrating the aggregate result for the relative strength of all capital types across baseline, weak, and strong capital projections for each of the 77 major suburbs in the City of Cape Town. Appendix C presents both the baseline capital and exposure scores classified from high to medium to low.

Figure 2. Baseline and medium-term future capital scores by major suburb in the City of Cape Town: a.) Strong Capitol Scenario, b.) Baseline Capitol Scenario, c.) Weak Capitol Scenario.

A broad pattern occurs of higher baseline capital scores across the Western and Atlantic seaboards from Cape Farms North down to Cape Point. In contrast, the central and eastern areas of the City of Cape Town display generally lower capital scores. In future projections of weak and strong growth in the various capitals, the rank order of suburbs remains largely the same over time.

It is interesting to note which capitals score the highest in the various suburbs as depicted in Table 3. Much variation occurs between the types of capital that dominate in the various suburbs, highlighting the value of using a holistic six capital framework in the assessment. However, there is no spatial differentiation in financial capital scores across the top 28 major suburbs as budget allocations are made at a much higher level than individual suburbs.

Table 3. Top five baseline capital scores per suburb.

Score rank	Human	Natural	Social	Manufactured	Intellectual	Financial
1	Langa (0,5)	Table Mountain (0,85)	Simons Town (0,8)	Observatory (0,72)	Table View (0,82)	Cape Farms North (0,66)
2	Delft (0,35)	Signall Hill/Lion’s Head (0,82)	Melkbosch Strand (0,56)	Cape Town (0,61)	Blouberg (0,55)	Brooklyn (0,66)
3	Gugulethu (0,33)	Cape Farms South (0,79)	Cape Point (0,32)	Sea Point (0,61)	Plattekloof (0,5)	Observatory (0,66)
4	Blue Downs (0,30)	Cape Point (0,77)	Cape Farms North (0,17)	Pinelands (0,51)	Rondebosch (0,43)	Blouberg (0,66)
5	Kalksteenfontein (0,30)	Camp’s Bay (0,7)	Durbanville (0,16)	Paarden Eiland (0,5)	Newlands (0,42)	Signal Hill/Lions Head (0,66)

For the full listing for each of the capitals across the 77 major suburbs, see the supplementary material.

Climate exposure assessment

Different capital stocks throughout the City of Cape Town are exposed to relatively different levels of climate stressors and shocks in space and over time. Figure 3 depicts relative levels of exposure to climate change for the baseline and medium-term future projections, respectively. It is important to note that the variance in climate exposure is relatively small in absolute terms across the City of Cape Town. As with the aggregate capital scores, exposure is a function of multiple metrics, which influence aggregate levels differently (Petrie et al. 2019). Broadly, the southern peninsula is less exposed, while the central and eastern areas are more exposed.

Figure 3. Baseline and medium-term future climate exposure scores by major suburb in the City of Cape Town: a.) Exposure (High), b.) Exposure (Baseline), c.) Exposure (Low).

Note: The colour spectrum used to represent data was chosen to clearly illustrate small discrepancies in climate exposure.

Capital-at-risk

Capital-at-risk combined relative exposure and capital scores in one aggregate score (aggregate data and descriptive statistics are included in the supplementary material). The results for various scenarios are displayed in Figure 4, presenting the relative spatial distribution of four scenarios for combined capital and climate exposure scores. Highest capital-at-risk scores are on the Western and Atlantic seaboards from Cape Farms North down to Hout Bay and Simon’s Town as well as Gordon’s Bay in the east. In contrast, the central eastern areas of the City of Cape Town display generally lower capital-at-risk scores. These scenarios represent the best and worst medium-term future positions based on currently available data. Intuitively, scenario 3 (weak capital growth and a low exposure to climate change) leads to the least risk to capital across the city, while scenario 4 (strong capital growth and a high exposure to climate change) leads to the greatest risk to capital.(Appendix C provides further detail on the baseline capital scores and exposure across each of the major suburbs in the City of Cape Town.)

Figure 4. Capital-at-risk across capital and exposure scenarios.

When explicitly disaggregating on a major suburb level, Figure 5 presents capital-at-risk as a scatter plot of baseline capital scores and baseline levels of exposure to climate change. The size of the circles indicates relative population densities in each of the 77 major suburbs colour-coded for each of the eight planning districts in the City of Cape Town. There is a clear general trend across the city showing areas of higher population (depicted by relatively larger circles) correlating with lower capital scores and medium to higher levels of exposure.

Figure 5. Capital-at-risk by major suburb in the City of Cape Town.

Results and discussion

Accurate quantitative economic assessments of urban-scale climate risk in spatially heterogeneous settings remain highly challenging. Our aim was to develop a practical and spatially sensitive approach to economic risk assessment of climate change at the city level that informs decision-making on climate responses in spatially heterogeneous cities – with specific reference to the City of Cape Town. We believe that the approach presented here is conceptually simple, flexible, and broadly implementable if (i) climate exposure is understood at the city level and (ii) if spatial datasets on selected leading or composite indicators for the identified capitals are available across city delineations at sufficiently high resolution – in our case a listing of major suburbs and key planning districts. Assessing the economic risks of climate change over time is essential for an effective climate response strategy. The capital-at-risk analysis presented here is designed to provide a spatially robust departure point for more detailed assessments on how economic risks of climate change may manifest across space and time. The approach presented is based on best available down-scaled climate modelling evidence of climate exposure at the City of Cape Town urban level (Petrie et al. 2019), and the results have already been used as input to the formulation of the City of Cape Town’s climate response strategy (City of Cape Town 2020b). The spatially robust capital-at-risk approach presented in this paper includes a broader set of socio-economic risks at a more refined spatial scale (to the suburb level) than what integrated assessment models used to estimate economic impacts and risks would typically allow for (see for example Estrada and Botzen 2021). Here, we discuss the main results and highlight some of the limitations of our approach.

Spatially, the suburbs with the highest levels of capital generally have high levels of financial capital, but with much variation in the levels of manufactured, social, natural, human, and intellectual capital (Table 3 and supplementary material). The suburbs with the highest baseline capital score are generally on the Western seaboard (geographically from Cape Farms North across Table View and Cape Town to Cape Point in the south). Manufactured capital scores are highest in areas all along this geographic area, namely Observatory, Cape Town CBD, Sea Point, Pinelands, and Paarden Eiland. However, when the various capitals are analysed further several notable exceptions to this trend do appear. The suburb of Langa has the highest human capital (number of employed people) of all the suburbs but relatively low levels of all other types of capital (except financial capital reflecting high levels of municipal budget expenditure in this suburb). Delft, Gugulethu, Blue Downs, and Kalksteenfontein are all suburbs geographically placed towards the centre of the City of Cape Town and score low on overall capital scores but are in the top five list of human capital in the City of Cape Town. The highest natural capital scores in the city are to be found in suburbs spanning across the City of Cape Town’s iconic Table Mountain to Cape Point mountain range as well as on Cape Farms South, but also to the east at Stellenbosch Farms and Gordon’s Bay. The highest intellectual capital in the city is in Table View and Blouberg, followed by areas such as Plattekloof in Tygerberg and Rondebosch and Newlands in the Southern suburbs. Social capital tends to be relatively low in the city according to our measure, with Simon’s Town and Melkbosch Strand and to some extent Cape Point as notable exceptions. What these observations reveal is that capitals other than financial or manufactured capital are often distributed spatially very differently.

The results on capital-at-risk estimates on a major suburb level clearly indicate where currently the highest aggregate capital scores overlay with the highest relative climate exposure (upper right area in Figure 5). From an economic risk perspective, these are the areas where capital-at-risk is the highest and which would accordingly inform a spatially sensitive climate response strategy. These areas coincide strongest with the Blaauwberg, Table Bay, and Northern planning districts. As we have conceptualised a holistic vision of various capitals, it is insightful to note that certain areas with very high human capital (e.g. Langa) and certain areas with very high natural capital (e.g. Table Mountain) also come out as areas with medium to high capital-at-risk to climate exposure. As these examples illustrate, investing in the productive value of the economy to minimise risks of climate exposure would not only focus on protecting and enhancing infrastructure and property values but would also include investing in the workforce and in city’s natural assets.

Another important result illustrated in Figure 5 is that a people-centred approach to risk management would have to recognise not only the productive value of the economy (as measured through the capitals) but also the population at risk to climate shocks and stressors. In the spatially heterogeneous City of Cape Town, the results indicate that areas with relatively higher populations (indicated by larger circles) are correlated with lower overall capital scores and medium to high climate exposure (lower half in Figure 5). These areas coincide strongest with the Cape Flats, Tygerberg, and Mitchell’s Plain/Khayelitsha planning districts. The pattern is largely the result of South Africa’s history of population separation, which is still evident in population distributions and settlement patterns today. These areas have attracted relatively less investment in infrastructure, are often not formally planned, and do not have access to much social or natural capital. Although government investment in these areas has been increasing over time (as measured by the financial capital indicator), these areas remain with infrastructure deficits and attract relatively little private investment. These areas also suffer from relatively higher levels of crime. Not surprisingly, many of these areas also exhibit low levels of overall resilience to climate change and other crises (Petrie et al. 2019). Increasing the resilience of a population to climate change is an important part of any climate response strategy, but on a practical level it raises issues for climate adaptation strategies aiming to be developmental and inclusive (City of Cape Town 2020b). Our approach to view capital holistically, rather than narrowly focussed on financial or manufactured capital, is an attempt to ameliorate the gap between managing the risks to a productive economy and improving the resilience of the people. The ‘right kind of growth’ and development is needed, namely an inclusive and resilient growth strategy that reduces the risks of and vulnerability to climate change (Bowen et al. 2012). For example, investments in property and infrastructure do need to recognise climate risks, as well as the viability of ecosystem-based adaptations to climate change (Cartwright et al. 2013; Prober et al. 2015). In a recent review on climate change in South Africa, Ziervogel et al. (2022) points out that urban ecosystem-based adaptation efforts in African cities are ‘nearly all…experimental and supported by international rather than local finance’. This perspective was borne out in nine investigative pre-feasibility studies conducted across seven of South Africa’s metropolitan municipalities in 2022. This work was implemented by the World Bank and National Treasury and enabled by Swiss development cooperation. The studies explored a range of approaches to urban resilience, while each study highlighted the urgency for integrating climate resilience into urban development to optimise economic outcomes such as for land value capture¹ (World Bank forthcoming).

Although a spatial assessment as presented here gives more granularity for planning purposes than what the sectoral or economy-wide approaches of macro-economic assessments would allow, a main limitation of our approach is that it does not account for the myriad of interlinkages and interdependencies, which limit or reinforce the impacts of climate change throughout the economy. Climate change shocks such as floods, fires, and heatwaves present immediate impacts to several economic activities in sectors as diverse as agriculture, transport and water and electricity, and with multiple risk pathways into the rest of the economy. A city economy is not fragmented into spatial suburban units but operate as an integrated system with linkages to regional, national, and international economies. One of the implications is that it is not necessary for all suburbs to have equally high scores of all capital types as trade and exchange is a normal feature of wealth creation and risk management in a modern economy.

It is important to note that this analysis is designed to provide a departure point for more detailed assessments of how risks may manifest differently across space and time. For example, an area may exhibit high levels of capital-at-risk, but investigating the underlying data might reveal that the area is exposed only to expected increases in temperature, which could be ameliorated to a certain degree by investing in stronger natural capital through interventions such as ecological restoration (Prober et al. 2015; Bustamante et al. 2019). Certain climate-related risks would pose greater threats to different types of capital, for example fire would pose a significant risk to natural capital, and floods would pose a greater risk to manufactured capital such as infrastructure. More research is required in developing specific viable options to manage the economic risks of climate change in specific suburban contexts.

A key assumption of the economic risk assessment presented here is that the chosen leading or composite indicators are representative of the specific capital distributions throughout the city. Although the indicators were tested with City of Cape Town’s managers and the associated future scenarios are based on historical trends and scientific projections, the choice of indicator metrics was heavily influenced by available data at suitable spatial scales. Different cities may choose different indicators or suites of indicators. A further limitation of the assessment presented here is the choice of indicator variable for financial and social capital. Total budget expenditure only represents the financial capital available to the municipality and does not incorporate the productivity of the broader economy (e.g. including private sector financial capital). Certain areas are invested in more by the municipality than others; these are often areas that experience service delivery deficits and are generally lower-income areas. Thus, this may in some cases narrow the gap in aggregate scores between areas of higher capital and lower capital because higher investment in those areas is considered as a relatively higher financial capital score. Moreover, the inverse of the crime rate is a proxy for social capital in some respects, but this assumption is likely not to hold in all cases (e.g. higher rates of crime in some areas might result in improved social cohesion in response to the crime).

Conclusion

In conclusion, the practical and flexible approaches for economic risk assessment at a city level as presented here may especially hold value to city managers of spatially heterogeneous cities faced with climate change stressors and shocks now and in the foreseeable future. By adopting the six capital framework and applying it to data from the City of Cape Town across various climate exposure and capital growth futures, it was demonstrated on a major suburb level that investment in the productive value of the economy would include investment in areas that are relatively more exposed to climate change over time and which have relatively higher levels of various capitals-at-risk: human capital (e.g. Langa and Delft), natural capital (e.g. Table Mountain, Signal Hill/Lion’s Head, Cape Farms South, and Cape Point), social capital (e.g. Simon’s Town and Melkbosch) and intellectual capital (Table View, Blouberg, and Plattekloof). Understanding the economic risks of climate change is essential to effectively adapt investment planning. The capital-at-risk analysis is designed to provide a robust departure point for more detailed assessments of how risks may manifest differently across space and time.

The assessment provided in this paper presents a framework to assist policy development and investment planning to accurately target the implementation of adaptation options through the understanding of relative levels of capital endowment vis-à-vis relative levels of exposure and through various climate-related hazards. Other practical use cases for this research include but are not limited to the development of climate change action plans, urban resilience strategies, infrastructure development plans, land value capture strategies, and spatial planning efforts, which will all influence city budget allocations.

The expanded definition of capital introduced in our analysis allows for evidence-based support of a growth and development strategy focussed not only on productivity (as usually measured through financial metrics) but also inclusiveness and the resilience of people to climate change. Our results could be used to support a shift in urban spending towards the protection of all capital categories broadly defined (financial, manufactured, human, social, intellectual, and natural), in the areas most exposed to climate change.

Acknowledgments

The authors would like to acknowledge Amy Davison of the City of Cape Town (Head – Climate Change – Environmental Management Department) for her immeasurable support of this project and research, including overall project coordination, facilitating data access and stakeholder engagement, and verifying and validating the findings and outputs.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This research was funded by the Agence Française de Développement (AFD) for the benefit of the City of Cape Town, grant number AFD: CZZ2197/MS/2017/03, under the project Accord-Cadre de prestations d’études et d’assistance technique pour l’initiative Villes et Changement climatique en Afrique subsaharienne (CICLIA), AFD/DOE/EBC/CLD | ACH-2017-026

Notes on contributors

Martin de Wit

Martin de Wit is Professor: Environmental Governance at Stellenbosch University. He received his DCom degree at the University of Pretoria in 2001. His research is organised around an economic approach to environmental management and policy, and follows an interdisciplinary approach with other experts in the fields of climate change, biodiversity and ecosystem services, waste management, energy transitions and ecological restoration. Martin serves as a regional coordinator for the African Association of Environmental and Resource Economists (AfAERE).

Jonty Rawlins

Jonty Rawlins is a sustainable development consultant and director at Sustainable Development Africa. He received an MSc in Water Science, Policy and Management from Oxford University in 2017 and an MSc in Economics from Rhodes University in 2016. His work and research are focusses on practical and theoretical approaches to sustainable development and environmental management, with a particular focus on climate change, water resources management, biodiversity conservation, ecosystem services and equitable socio-economic development.

Belynda Petrie

Belynda Petrie is co-founder and CEO of OneWorld Sustainable Investments and a PhD candidate at the University of Cape Town, where she is developing a thesis on regional integration and transboundary water governance and security. Her work spans policy and strategy developing in climate change response, water and energy security, and urban resilience across Africa and Asia. She serves as a member of the Science Programme Committee for World Water Week and advises on climate finance on various boards, including the Alliance for Global Water Adaptation, and the recently launched Continental Africa Water Investment Programme.

Notes

1. Land value capture is a policy approach rooted in the idea that public action should result in public benefit. As such, this approach seeks to enable communities to recover and reinvest increase in land value that result from public investments and government action. South Africa’s Land Value Capture Programme was launched in 2020 through a tripartite partnership between the Development Action Group, the Lincoln Institute of Land Policy, and the National Treasury’s Cities Support Programme.

Appendices Appendix A:

Indicators for the six types of capital

Table

Capital type	Indicator	Data/measurement
Financial	Operating cost (actual/budget) of city	Service charges, property rates, government grants, other own revenue, investment revenue
	Capital costs (actual/budget) of city	Borrowing, internally generated funds, national grants, provincial grants, public donations
Manufactured	Areas and property values over various built environment categories (e.g. commercial, industry, residential)	Median freehold residential property values, properties by suburb (770 units) and value band
	Roads, bus-roads, railways	Kilometres of roads, bus-roads, railways
	Vehicles over various categories	Number of households with motorcars
	Critical infrastructure	Proxy of travel time to critical infrastructure and service centres (e.g. fire departments, police stations)
Human	Economically active population	Number employed, unemployed, discouraged and non-active persons
	Indicators of (un)employment, poverty and inequality	Number (un)employed, median household income, income range by suburb, households with cellphone/internet access
	Educational attainment across age groups	Percentage of population with no schooling (aged 20+), with matric (20+), with higher education (20+)
	Health indices	Pneumonia incidence and malnutrition incidence (under 5), hepatitis A incidence, typhoid incidence, HIV incidence, percentage positive TB tests
	Informal settlements	Percentage area of informality
Intellectual	Size of knowledge sector	Number of people with higher education
	Money spent on research and development	R&D as percentage of GDP
	Brand awareness	Brand perception
	Science outputs	Number of publications
	Number and value of knowledge, technology patents	Number of patents
Natural	Area of land across natural categories	Area of land declared urban conservation areas, as nature reserves, as urban conservation areas, designated as wetland
	Water stocks and use	Average monthly household water use
	Measures of biodiversity integrity and resilience	Percentage of suburb area that are critical biodiversity and protected areas
	Pollution metrics	Water, air and solid pollution
	Waste metrics	Percentage of households with no rubbish collection
	Ecosystem services metrics	SANBI ecosystem status (mean)
	Valuation of environmental quality of air, land, water, biological systems	Value of ecosystem services
Social	Measures of trust, reciprocity and cohesion	Racial integration index
	Indicators of livelihoods and dependency	Household dependency ratio Number of street people
	Metrics of involvement in social and cultural initiatives	Number of early childhood development forums
	Metrics of citizen satisfaction	Percentage of people who voted for governing party Number of service requests lodged
	Metrics of social wellbeing	Access to piped water, electricity Housing owned/paid off Population density Incidences of crime

Appendix B:

Locality of thetempmajor suburbs in the City of Cape Town

Appendix C:

Capital scores and exposure across 77 major suburbs in the City of Cape Town

Further detail on the relative baseline capital scores and exposure across major suburbs in the City of Cape Town is provided here. The scores are determined as follows:

• High>75th percentile

• Medium<75th percentile & >25th percentile

• Low<25th percentile

Table

Major suburb	Baseline capital	Baseline exposure
AIRPORT	Low	Medium
ATHLONE	Low	Medium
ATLANTIS	Medium	Medium
BELLVILLE	Low	Medium
BERGVLIET	Medium	Medium
BISHOP LAVIS	Low	Medium
BLACKHEATH	Low	Low
BLOUBERG	High	Low
BLUE DOWNS	Medium	Low
BRACKENFELL	Medium	Low
BROOKLYN	Medium	High
CAMPS BAY	High	Low
CAPE FARMS NORTH	High	Medium
CAPE FARMS SOUTH	Medium	High
CAPE POINT	Medium	Low
CAPE TOWN	High	Medium
CONSTANTIA	Medium	Medium
DELFT	Low	Medium
DURBANVILLE	High	Medium
EERSTE RIVER	Low	Medium
ELSIES RIVER	Low	Medium
EPPING	Low	High
EVERSDAL	Medium	Medium
FALSE BAY COASTAL PARK	Medium	Medium
FISH HOEK	Medium	Medium
GOODWOOD	Medium	Medium
GORDONS BAY	Medium	High
GRASSY PARK	Medium	High
GREEN POINT	High	Low
GUGULETU	Low	Medium
HANOVER PARK	Low	Medium
HOUT BAY	High	Low
JOOSTENBERG VLAKTE	Medium	Medium
KALK BAY	Medium	Low
KALKSTEENFONTEIN	Low	Medium
KENRIDGE	Medium	Medium
KHAYELITSHA	Low	Low
KOMMETJIE	Medium	Low
KRAAIFONTEIN	Medium	Medium
KUILS RIVER	Low	Low
LANGA	High	High
MACASSAR	Medium	Medium
MAITLAND	Medium	High
MALMESBURY FARMS	Medium	Medium
MAMRE	Medium	Low
MANENBERG	Low	Medium
MELKBOSCH STRAND	High	Low
MILNERTON	High	High
MITCHELLS PLAIN	Medium	Low
MUIZENBERG	Medium	High
NEWLANDS	Medium	High
NOORDHOEK	Medium	High
OBSERVATORY	High	High
OCEAN VIEW	Medium	Medium
OTTERY	Low	Medium
PAARDEN EILAND	Medium	High
PAROW	Low	Medium
PELIKAN PARK	Medium	Low
PHILIPPI	Low	Low
PINELANDS	High	High
PLATTEKLOOF	High	Medium
PLUMSTEAD	Medium	Medium
RETREAT	Low	High
RONDEBOSCH	High	High
SEA POINT	High	Low
SIGNAL HILL/LIONS HEAD	High	Low
SIMONS TOWN	High	Low
SIR LOWRY’S PASS	Medium	High
SOMERSET WEST	Medium	High
STELLENBOSCH FARMS	Medium	High
STRAND	Low	High
TABLE MOUNTAIN	High	Medium
TABLE VIEW	High	Medium
TOKAI	Medium	Medium
VREDEKLOOF	Medium	Medium
WELGEMOED	High	Medium
WYNBERG	Medium	Medium