Addressing Climate Change in Cyprus: Policy Options and Strategies
Rising temperatures, a likely intensification of extreme weather, and sea-level rise pose macroeconomic risks that must be closely monitored. Slow-moving warming is expected to reduce GDP per capita and sea-level rise can cause permanent annual economic and welfare losses. This calls for effective and efficient adaptation integrated into development planning. In the case of sea-level rise, building coastal protection is highly cost-effective, but a balanced mix of protection and planned retreat from the coastline is likely the least-cost strategy.
A. Summary
1. Warming, droughts, heavy-rainfall events, and sea-level rise already affect Cyprus and will likely intensify in the future. The warming trend observed from the 1950s is expected to continue even with strong mitigation efforts globally. It is also certain that sea level rise will persist, posing a significant threat to key island infrastructure situated along the coast, including airports, ports, power and desalination plants, tourist resorts, and other critical facilities. These effects can lead to an economic cost, increased vulnerability to storm surges, and challenges in maintaining the functionality and sustainability of coastal regions.
2. Adaptation can be very effective at reducing the cost of sea-level rise, but it requires careful planning and a balanced mix of protection and planned retreat. IMF staff estimates and other studies indicate that SLR can cause costs up to 0.4 percent of GDP annually after 2050 without adaptation efforts. An adaptation strategy that combines protection and planned retreat of areas at risk of inundation would cut the overall cost of sea-level rise (including adaptation costs) by 70 to 90 percent to approximately 0.1 percent of annual GDP by mid-century. The cost of other adaptations can be contained by focusing on cost-effective measures and delaying adaptations that will be cost-effective only in worst case climate outcomes.
3. To be effective and efficient, adaptation to climate change must be an integral part of development planning. With many competing needs, the government must carefully allocate resources across all possible uses, including adaptation to climate change, while considering the distributional effects of its programs. This requires: (i) concentrating government efforts and resources in key areas; and (ii) collecting information on the effectiveness of spending across alternative programs and on how spending affects distinct groups in society (Bellon and Massetti, 2022a). The government can prioritize adaptation policies with positive externalities, by removing market imperfections and policies that hinder efficient private adaptation and by ensuring a just transition.
B. Overview of Climate Trends and Projections
4. Cyprus experiences a pronounced Mediterranean climate characterized by distinct seasonal patterns in temperature, rainfall, and overall weather. In the Eastern Mediterranean and Middle East (EMME) region, Cyprus has a Mediterranean climate with average annual temperatures ranging from 16°C to 19°C and a mean of 17.5°C. The warmest year recorded was 2010 at 20.4°C, and annual precipitation varies, reaching a low of 200 mm in 1972–73 and a high of 800 mm in 1968–69, with a mean of 503 mm for 1961–90. The EMME is identified as a significant climate change hotspot. The region has experienced faster warming than the global average and other inhabited parts of the world areas, leading to noticeable changes in the hydrological cycle, including more severe and prolonged heatwaves, droughts, dust storms, and flash floods (Zittis et al. 2022).
5. Cyprus average yearly temperature in both urban and rural areas is on the rise. Over the past decade, the majority of Cyprus, particularly the three major cities, has faced elevated temperatures, causing discomfort and significant challenges, including increased energy consumption for cooling, higher water usage, and a heightened risk of forest fires. Urban areas, impacted by urbanization1, show a significant temperature rise, while rural areas also indicate a broader regional and global warming trend with their observed temperature increase (UNFCC for Cyprus).
6. Cyprus has experienced a noticeable surge in hot days. With temperatures steadily rising, summers have become increasingly intense, characterized by prolonged periods of extreme heat and more frequent heatwaves (Figures 1 and 2). This uptick in extreme heat poses significant challenges to the island nation, in particular for agriculture, public health, and infrastructure. In Nicosia, the average annual temperature rose from 18.9°C to 20.9°C over the past 30 years, signifying a 2.0°C increase and an additional 20 days per year under extreme heat2 (UNFCCC of Cyprus).
Annual Change in Hot Days and Annual Average Number of Hot Days
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A003
Note: Hot days are defined as days when the daily maximum temperature (TX35) exceeds 35°C. Change in the map represent the difference between the 30-year averages (1991–2020) and (1961–1990) for the specified variable. Gridpoints marked with dots indicate locations where the changes are not statistically significant. The time series on the left panel displays the national average number of hot days.7. With low and volatile rainfall, Cyprus is vulnerable to a small decline in precipitation. Observations highlight a notable centennial scale decline in the annual distribution of rainfall and changes in interannual variability during the 20th century. Since 1950, average total annual precipitation has remained nearly unchanged, but volatility is large. In this semi-arid Mediterranean climate, the island’s water resources are vulnerable to frequent and prolonged periods of drought 3. A reduction of total rainfall caused by climate change could worsen the already severe pressure currently experienced by the water sector.
8. Cyprus’s temperatures are projected to continue to increase for decades. In the three IPCC emission scenarios4 that cover both uncertainty in future emissions (scenario uncertainty), and uncertainty in the response of the climate, estimates suggest Cyprus will experience additional warming ranging from 1.1°C to 1.6°C by the end of the century relative to the 1995–2014 baseline, with a potential increase of up to 2.5°C in the high-emission SSP3–7.0 scenario. Cyprus is anticipated to undergo less extreme temperature increases compared to some other countries in the EMME. However, starting from baseline temperatures relatively high compared to most advanced economies, the island faces risks associated with heat stress that necessitate ongoing monitoring. The consensus among model projections is that total annual precipitations will decline using all emission scenarios, but there is large uncertainty on the magnitude on projected changes.
9. Forest fires are a present threat but establishing robust trends in forest fires and attributing them to climate change are difficult. Forest fires in Cyprus present a threat to both ecosystems and residential areas, especially in periods with high temperatures, prolonged droughts, strong winds, and in presence of flammable vegetation (Alker, 2009). The 2021 record-breaking wildfire in Cyprus, the largest in decades, was exacerbated by a week-long heatwave with temperatures surpassing 40°C. However, establishing robust trends in forest fires and attributing them to climate change are difficult because of many concurring factors. Despite weather conditions that are favorable to forest fires becoming more common, forest fire statistics from 2000–2021 indicate a significant 50% reduction in number of fires and burnt area, potentially thanks to successful management measures implemented by the Department of Forests. This is an encouraging dynamic because weather conducive to forest fires, assessed through the Fire Weather Index (FWI) and a regional climate model (Kostopoulou et al., 2014), is expected to become more common. Both high fire risk days (5–15 days/year) and extreme fire risk days (1–10 days/year) are projected to increase in 2021–2050 compared to 1961–1990.
Average Annual Temperature and Total Annual Precipitation
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A003
Note: SSP1–2.6 is in line with the Paris goal to keep global mean temperature increase below 2 ºC with respect to pre-industrial times. SSP2–4.5 represents continuation of present trends. SSP3–7.0 is a high emission scenario.10. It is certain that the sea level will continue to rise and that Cyprus faces an increased risk of coastal erosion, loss of land, and potential damage to infrastructure and communities. Figure 3 displays median projections of sea-level rise (SLR) using different emission scenarios from a leading study (Kopp et al. 2014). In Cyprus, the sea level is projected to increase by 0.4 to 0.7 m in 2090–2099 with respect to 2000–2009 (approximately 0.3 to 0.6 m above the present level) depending on the emission scenario.5 SLR is projected to be lower than the global mean, a result confirmed by other projections (Slangen et al., 2014). While there is no doubt that the sea level will continue to increase for ongoing decades (achieving the Paris goals slows down but does not stop SLR at least until the end of the century), projections are uncertain due to uncertain emission trajectories and uncertain response of the sea level to different warming rates. The right panel of Figure 3 illustrates the scientific uncertainty about sea-level rise for a moderate emission trajectory (RCP 4.5): the median projection is that sea level will rise by 0.6 m with respect to 2000–2009, but SLR in the range 0.30 to 0.95 m cannot be excluded.
Sea-level Rise Scenarios for Cyprus
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A003
Note: Left panel: Local and Global Sea-Level Rise (SLR) median projections until 2100 under three emission scenarios: RCP2.6 (Paris 2° C target), RCP4.5 (Moderate), and RCP8.5 (Extreme). Solid lines depict median local SLR for each emission scenario. Dotted lines depict the median global SLR. Right panel: Local SLR probabilistic projections until 2100 using the RCP4.5 scenario. Solid line depicts the median projection and dotted lines the 5th and 95th percentiles of the distribution.C. Macro-Economic Risks
Slow-Moving Warming and Extreme Weather
11. IMF staff analysis and a review of the literature identify moderate risks from climate change in Cyprus with large uncertainties. Slow-moving warming and sea-level rise are expected to be a drag on the economy, especially without adaptation measures. There is no evidence of significant macro-economic impacts from weather disasters both using historical data and projections, but substantial uncertainty remains about future changes in climate and their potential economic impacts. Large negative shocks from abrupt climatic and environmental changes cannot be excluded but they are impossible to quantify.
12. Several lines of evidence indicate that the cumulative effect of slow-moving warming can reduce real per capita GDP by 3 to 5 percent by 2100 in a fast-warming planet. This range is calculated using results from two econometric models and estimates from a European research project (Box 1). This warming scenario accounts for larger-than-expected warming rates (90th percentile of the temperature range projected by climate models using the high-emission SSP3–7.0 scenario), but effects of changes in extreme weather and sea-level rise are not included.
13. Climatological disasters do not currently pose severe macro-economic risks, and risks are not projected to increase substantially, but large uncertainties remain. Heat waves present the main present risks for the population and are expected to become more intense (Ministry of Agriculture, Rural Development and Environment 2023; Chapter 2, Section 6.3.4.4). Heat waves in 1998, 2000, and 2022 caused an estimated total number of 158 deaths and in 2000 high temperatures have been associated with drought and forest fires (Figure 4), but the overall economic impact has been limited and cannot be easily assessed. Projected climate change impacts on infrastructure are limited (Ministry of Agriculture, Rural Development and Environment 2023, Chapter 2, Section 6.3.2). Flash floods pose risks, but they are linked more to land use changes than to changes in rainfall intensity. Advanced econometric analysis that relies on machine learning methods and very granular weather data does not reveal significant macro-economic shocks from a multitude of weather extremes in the present (Box 1). Projections by the Network for Greening the Financial System (NGFS) do not indicate significant changes in annual expected damages from river floods, and changes in other hazards are projected to be small.6 There are however large uncertainties in how physical hazards will evolve due to the inability of climate models to predict with accuracy local events over short time scales, like storms. There is also uncertainty in potential economic, social, and environmental impacts of these events. The expected continuation and possible intensification of dry conditions represents a downside risk with potentially large yet unknown impacts in an already water-scarce country. This suggests monitoring risks and preparing contingency plans also for low-probability events.
Count of Climatological Natural Disasters According to the EM-DAT Database
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A003
Source: EM-D AT , CRED.Macro-Economic Impacts of Gradual Warming Kahn et al. (2021)
IMF staff has used estimates of the impact of warming on GDP growth from Kahn et al. (2021) to estimate the long-term effect of gradual warming. Kahn et al. (2021) connect “deviations” of temperature and precipitation (that is, weather) from their long-term moving-average historical norms (that is, climate) to growth in real GDP per capita. This theoretical model is then estimated by using data from 174 countries over 1960–2014.
Per-capita real output growth is affected by persistent changes in the temperature trend. An acceleration of the warming trend relative to historical values, has a negative impact on GDP forecasts while a deceleration of the trend has a positive impact. Abiding by the Paris Agreement goals (SSP1–2.6), thereby slowing the warming trend with respect to historical values, raises projected GDP above reference. The SSP2–4.5 scenario is close to continuation of present trends and has generally small negative or positive impacts. The SSP3–7.0 scenario represents an acceleration of the present trend and generates negative impacts. The 90th percentile of the SSP3–7.0 scenario provides a worst-case outcome and generates the largest negative impacts.
Adaptation is implicitly included in the model, without accounting for its cost. After 30 years countries are assumed to have fully adapted to the observed trend. If the trend persists along the lines observed during the previous 30 years, there is no impact on projected GDP.
Network for Greening the Financial System (NGFS)
The NGFS estimates losses of GDP per capita using empirical work by Kalkhul and Wenz (2020). Using RCP8.5 warming rates, which are similar to warming rates in the SSP3–7.0 High scenario, the NGFS estimates a loss of GDP per capita equal to 7 percent in 2100.1 This loss is measured against a hypothetical scenario of no climate change while losses in Kahn et al. (2021) are measured against a scenario in which temperature continues growing following historical trends. Comparing the elasticity of GDP losses to changes in warming rates reveals that estimates based on Kahn et al. (2021) are approximately 2 percentage points lower than those based on Kalkhul and Wenz (2020), which is a minor difference considering large uncertainties in very long-run projections.
SOCLIMPACT
SOCLIMPACT is a research project funded by the European Union that assessed climate change impacts and their socio-economic implications for European islands, including Cyprus (https://soclimpact.net). Estimated impacts of temperature rise, fire risk, beach erosion, and sea-level rise on tourism, energy demand, and port infrastructure are aggregated at the macro-level using the computable general equilibrium model GEM-ER-ISL. Results show that with the RCP 8.5 scenario (analogous to the SSP3–7.0 High scenario) impacts are in the 3 to 5 percent range (SOCLIMPACT, 2020; pp. 64–81). Another CGE, GINFORS, reports results only up to 2050 that are within the 3 to 5 percent range but could become larger in the second half of the century.
Akyapi, Bellon, and Massetti (2022)
Akyapi, Bellon, and Massetti (2022) (ABM) use hundreds of billions of high-resolution and high-frequency weather observations to build hundreds of weather variables that can potentially affect macroeconomic variables.
Macroeconomic Impact of Gradual Warming
(Percent change of GDP)
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A003
Source: Centorrino, Massetti, and Tagklis (2023).To select a limited and meaningful set of variables ABM use machine learning algorithms and then standard regression methods to estimate the macroeconomic effect of the selected weather variables. A global analysis that relies on a panel of all countries finds that severe droughts and days with maximum temperature above 35°C significantly reduce GDP per capita growth and have complex fiscal implications. An increase in the frequency of days with moderate temperatures between 9 and 12 C significantly increases GDP per capita. These variables do not have a significant effect on GDP growth of Cyprus. As other variables may be important in Cyprus, variable-selection algorithms are re-applied using only Cyprus macroeconomic time series but just one or two weather variables are selected as being relevant, and they do not have a statistically significant effect on GDP growth, government expenditure, or government revenue. This suggests that the economy of Cyprus is resilient to weather shocks and although weather extremes can cause damages, discomfort, and loss of life, their effect is not clearly distinguishable at macroeconomic level.
Source: CLIMADA, Climate Impact Explorer of Climate Analytics for the NGFS. 1 Cyprus page, consulted on 3/7/2024.Sea-Level Rise
14. Cyprus cannot control global sea-level, but it can manage how it affects the country by adapting. Even if efforts to keep global temperature increase in line with the Paris Agreement goals, the sea level will continue increasing throughout this century and beyond (Figure 2). Mitigation policy is key to limiting the speed of sea-level rise (SLR) during this century and the overall extent of SLR in future centuries, but the only practical way to limit impacts of SLR this century is by adapting.
15. IMF staff estimates, and other studies indicate that SLR can cause damages between 0.1 and 0.4 percent of GDP annually after 2050 assuming a moderate emission scenario, with large uncertainties. IMF staff estimates using the CIAM model are on the lower end of this range and are shown in detail in Box 2. The EU-funded project PESETA IV estimates costs on the higher end of this range due to larger storm-surge projections (Vousdoukas et al. 2020, Table 2). With an extreme projection of SLR, PESETA IV estimates costs of up to 4.9 percent of GDP in 2100, a very large potential loss (Vousdoukas et al. 2020, Table 2). These differences reflect large uncertainties in estimating SLR impacts.
16. Adaptation can greatly contain SLR losses at modest cost, but the optimal mix of protection and planned retreat remains uncertain and deserves more granular analysis. The CIAM model indicates that protection is highly effective at reducing losses from permanent inundation of coastal land, but planned retreat from the coastline is the cheapest option (Box 2). The EU project PESETA IV is more pessimistic on the cost of SLR without adaptation, but more optimistic on the cost and effectiveness of coastal protections. A modest investment of €7–10 million per year (approximately 0.04 percent of GDP) can reduce SLR costs by approximately 90 percent (Vousdoukas et al. 2020, Table 5). SOCLIMPACT, another EU-funded study, finds that upgrading seaports to keep them operable with SLR costs €2.2 million per year in 2046–2065 and €4.3 million per year in 2081–2100 when using an extreme emission scenario (SOCLIMPACT Deliverable Report -D5.6). While the optimal mix of protection and planned retreat remains to be determined case-by-case using high-resolution analysis—like the ongoing COASTANCE and MAREMED projects (Ministry of Agriculture, Rural Development and Environment, 2023, p. 220), there is universal agreement that a forward-looking adaptation strategy can be highly effective at reducing SLR impacts with relatively low costs.
17. Cost-benefit analysis (CBA) can be used to select efficient coastal adaptation strategies. CBA can be challenging, but even preliminary and incomplete assessments are useful to identify trade-offs and the most attractive policy options using a transparent and systematic approach (Bellon and Massetti, 2022a). Best practices can be drawn from coastal protection analysis and policies in other countries, for instance in the Netherlands, where there is a long-standing tradition of using CBA and cost-effectiveness analysis for flood risk management and water governance. This tradition started in 1954 with the pioneering CBA of the Delta Works by Tinbergen (1954) and continues to this day (CPB, 2017).
Estimating the Cost of Sea-Level Rise and Adaptation
The analysis of sea-level rise impacts, and adaptation options is done using complex models that rely on necessary simplifications but provide important insights. While there is uncertainty on the exact extent and cost of damages from SLR and on the cost of protection measures, there is consensus in this literature that long-term planning of adaptation can be highly effective at containing physical impacts and costs of SLR. For example, the large EU-funded research project PESETA IV finds that adaptation can reduce SLR damages in the EU by approximately 90 percent (Vousdoukas et al. 2020, Table 6) with an average benefit/cost ratio from 75 to 110 depending on the emission scenario. Model simulations fully agree that adaptation can be highly effective but may differ on the optimal mix of adaptation measures – e.g., hard protection, nature-based solutions, planned retreat – because they use different data, use different climate scenarios, or work under different normative criteria. There is also consensus that the transformations needed to adapt to SLR, while technologically feasible and economically sound, are complex and require strong governance (Hinkel et al., 2018).
IMF staff uses the state-of-the-art Coastal Impact and Adaptation Model (CIAM) to estimate the cost of sea-level rise under alternative adaptation strategies. CIAM is a global model used to estimate the economic cost and benefits of adaptation to sea-level rise (Diaz, 2016). The global coastline is divided into more than 12,000 segments of different length grouped by country. Cyprus’s coastline comprises 7 segments that are 44 Km long on average, ranging from 1 Km to 111 Km. This analysis does not cover areas of Cyprus not under the effective control of the Republic of Cyprus and assumes no change in status quo. Each segment is further divided into areas of different elevation. For each segment, the model has data on capital, population, and wetland coverage at different elevations. By using projections of local sea-level rise from Kopp et al. (2014), it is possible to estimate the areas that will be inundated and the amount of capital and population at risk. Storms cause periodic inundations on top of sea-level rise. The model does not consider increased risks from river floods. See Annex I for additional information.
The model calculates the cost of SLR—protection costs plus residual losses—under alternative adaptation options:
The no-adaptation scenario assumes that population does not move until the sea inundates the area and then relocates to areas with higher elevation. Society keeps building and maintaining capital until inundation causes irreversible losses and capital is abandoned. The cost of sea-level rise is calculated as Yohe et al. (1995), because as SLR progresses, coastal proximity rents will shift from land that is inundated to adjacent land. Population density and development opportunity costs are assumed to be capitalized in agricultural land values. The disutility cost of reactive migration is monetized and contributes to the cost of sea-level rise.
At the opposite, a protection scenario assumes that society invests in cost-effective seawalls and other barriers to avoid inundation from sea-level rise, but storms can still periodically inundate protected areas if protection is not sufficiently strong. Capital and land are not lost, the population does not move, but storms periodically cause capital and human losses. The cost of SLR is equal to the cost of protection plus the expected value of the cost of storms.
Another adaptation option relies on planned retreat from areas that will be subject to inundation. The goal of retreat is to keep using coastal areas without building new capital and by letting the existing capital depreciate. For example, a coastal road is used until it needs major retrofitting investment. Then, a new coastal road is built in-land on higher grounds. This strategy accepts that land and some residual value of capital will be lost, but it avoids coastal protection costs. The population gradually moves to higher grounds before areas are inundated. The cost of sea-level rise is equal to the sum of the residual cost of capital, the value of inundated land, and the disutility cost of migration.
Other adaptation strategies consider different degrees of protection against storms and different speeds of retreat. For each coastal segment, the model calculates the net present value of each adaptation strategy by summing discounted costs and benefits (avoided damages) over time. Loss of life is monetized using the Value of Statistical Life and loss of wetland is monetized using estimates of willingness to pay for biodiversity preservation.
The cost of building and maintaining seawalls, and other key parameters are from the literature. Storm surge costs are incremental with respect to a baseline scenario in which storms occur with lower sea-level. The optimal strategy is the strategy with the largest NPV and can differ across coastal segments within the same country. For example, protection is usually the optimal strategy in areas with large existing capital and high population density while retreat is optimal in areas with low capital and population density.
Despite many uncertainties and some necessary simplifying assumptions, the model provides a useful framework to systematically think about costs and benefits of alternative adaptation options to sea-level rise. More granular coastal modeling and more accurate mapping of assets can provide a more precise assessment of costs and benefits, but the key insights developed with a baseline version provide a useful starting point to deal with a complex, multidecadal challenge.
Cost of Sea Level Rise, 2020–2099
(Percent of GDP)
Citation: IMF Staff Country Reports 2024, 138; 10.5089/9798400276439.002.A003
Sources: Diaz (2016), Kopp et al. (2014), and IMF staff calculations.Note: Using a state-of-the-art model of sea-level rise costs and adaptation, IMF staff estimates the cost of local sea-level in Cyprus corresponding to the RCP 4.5 scenario in Kopp et al (2014) assuming three policy scenarios: (i) no planned adaptation – society reacts to sea-level rise by relocating, no protection is built and capital losses are large; (ii) Protection – society plans construction of cost-effective protection against sea-level rise by anticipating its effects without relocating people or assets; and (iii) Optimal Adaptation – society plans a mix of protection and retreat anticipating sea-level-rise, by comparing costs and benefits of each option and choosing the strategy with the largest net present value. Cumulative undiscounted costs divided by cumulative undiscounted GDP. Wiskers on top of each bar indicate the range of total cost using the 5th and 95th percentile of the probabilistic distribution of sea-level rise. Due to the highly non-linear nature of coastal impacts, adaptation costs, and effectiveness of adaptation measures, ranges are not always symmetric around total costs.D. Adaptation
18. Cyprus adopted a comprehensive National Adaptation Strategy in 2017 that identifies key intervention areas and specific adaptation actions. Starting from a vulnerability assessment, the Adaptation Strategy lists many adaptation measures in eleven priority sectors, among which water resources, coastline risks, biodiversity, and tourism have received priority. The implementation of several adaptation measures has already started (Ministry of Agriculture, Rural Development and Environment, 2023, p. 218).
19. The IMF has developed principles that can help select and prioritize adaptation actions amid other investment needs in an economic environment with growing public finance constraints. It is useful to think about adaptation in holistic terms, as a strategy to maximize long-term economic and social development that needs to become integral part of economic planning (Bellon and Massetti, 2022a,b; Aligishiev, Bellon, and Massetti, 2022). As such, investment and public resources diverted to adaptation need to be assessed against other potential uses (Bellon and Massetti, 2022a). It is then useful to identify priority areas for government intervention and tools to select and prioritize adaptation actions and investments.
20. With many competing needs, government efforts and spending would be most effective if concentrated on: (i) adaptation policies that are public goods; (ii) reforms that promote efficient private adaptation; and (iii) strengthening of social programs. Individuals and firms have strong incentives to adapt because many adaptation benefits tend to be local and private (Box 2). However, there is a clear role for government intervention when adaptation has large externalities, as in the case of coastal protection or strengthening of public infrastructure. As market inefficiencies and policy failures may limit private adaptation or create distortions, another key role for the government is to continue promoting reforms that foster the efficient use of all resources and ensure competitive access to markets (Bellon and Massetti, 2022a). Strengthening of social programs is likely needed to offset adverse effects of climate change and of adaptation policy itself on the most vulnerable part of the population. This will allow maximizing the impact of public spending and stimulate private adaptation, while ensuring a just transition and public support.
21. Despite limitations, cost-benefit analysis (CBA) can play an important role in helping decision makers to consistently collect, aggregate, and compare information on public adaptation projects. As exemplified by the analysis of sea-level rise in Section C, adaptation investment and policy will typically have trade-offs that would be better assessed by comparing social costs and benefits using a systematic approach. What to do, when, how, and at what cost ultimately rely on ethical choices that should reflect the preferences of society. However, cost-benefit analysis (CBA), complemented by analysis and correction of distributional impacts, can help decision makers maximize overall social welfare by avoiding wasting scarce resources. To achieve this goal, it is essential that CBA is applied to adaptation as well as to all other development programs in a consistent manner (Bellon and Massetti, 2022a). Limitations of CBA (e.g., lack of data, uncertainty in long-term projections, sensitivity to choice of discount rates) need to be carefully assessed. Cost-effectiveness analysis—the choice of the least-cost strategy to attain a desired level of protection—can be an alternative to CBA if outcomes are considered too uncertain by policy makers Bellon and Massetti (2022a).
22. According to standard CBA rules, only programs with NPV greater than zero should be financed. Competing programs should be ranked using CBA and only programs with the highest ranking should be financed. In the example of adaptation to sea-level rise, retreat is the strategy with the highest net present value. Similar comparisons can be developed when comparing alternative options to strengthen public infrastructure. By consistently investing in projects with the highest returns, governments can maximize the impact of their spending. This means, for example, saving the largest number of lives, providing high-quality public education, ensuring that social safety nets are well-funded, and boosting long-term growth (Bellon and Massetti, 2022a).
23. Compensation might be more efficient than investments in adaptation to achieve society’s equity preferences. Full protection of all assets and populations at risks may be very expensive in some cases, as shown in the case of adaptation to sea-level rise. As a result, adaptation projects may have a negative NPV (for example, protection against sea-level rise is more expensive than no action according to CIAM’s estimates). While there can be specific reasons to warrant investment even with a negative NPV, the authorities should consider if it is possible to support the affected population in alternative ways. This can take the form of relocation subsidies or other forms of supports with less stringent conditionality (Bellon and Massetti, 2022a). For example, compensation could be used to redistribute the cost of sea-level rise from coastal landowners to society if planned retreat is cheaper than full protection.
Market Reforms to Boost Adaptation in the Private Sector
In imperfectly competitive markets, adaptation is inefficient, and governments should intervene mirroring standard prescriptions for public policy from economic theory.
Some market imperfections pertain to the nature of the adaptation goods themselves. For example, markets invest sub-optimally in adaptations with large positive externalities and public goods, such as information about climate change, emergency preparedness plans, seawalls, basic research in new materials, and technologies to cope with higher temperature.
In many instances, resilience depends on networks, such as a system of dikes, a water network, or a transportation network. As adaptation in each part of a network has impacts on the rest of the network that may not be captured, private adaptation will tend to be underprovided. Government coordination may be needed to internalize all the benefits for society.
The extent of needed cooperation for adaptation projects depends on the extent of the externality that is addressed by the project. Building a more resilient storm water drainage system may only require cooperation at the city level. If risks from sea-level rise are localized, each locality may invest in its own system of protection. The central government can provide adaptations with local effects, but that would be equivalent to a transfer of wealth between regions when projects are financed from national resources. As risks grow in scope and complexity, cooperation might be needed at the national or even the international level, for example to manage floods in transnational rivers. In general, the optimal distribution of responsibilities across levels of government also depends on the existing allocation of responsibilities.
Other market imperfections affect the broad functioning of the economy and make adaptation to climate change inefficient. For example, a poor business environment and inefficient credit markets hamper opportunities for farmers to invest in new capital to grow crops that are more suitable to the new climate.
Moral hazard may cause insufficient investment in adaptation if consumers, firms, and local government expect central governments to provide relief. To avoid moral hazard, governments can implement regulations that minimize risk taking. Examples include zoning that prohibits construction in flood zones, building codes, mandatory evacuations, and mandatory insurance.
The government may also consider correcting market distortions resulting from their own policies (policy failure). For example, subsidies to inputs can lead to inefficient use. Of particular concern is subsidized water use, which may worsen water scarcity problems due to climate change. Barriers to international trade also prevent efficient climate-change-induced reallocation of capital, land use, and other resources to maximize their productivity. The government may consider removing these distortions as part of a comprehensive plan to improve the efficiency of the economy, while taking into consideration the distributional implications of these measures.
References
Aligishiev, Zamid, Matthieu Bellon, and Emanuele Massetti. 2022. “Macro-Fiscal Implications of Adaptation to Climate Change,” IMF Staff Climate Notes, No. 2022/002.
Akyapi, B., M. Bellon and E. Massetti. 2022. “Estimating Macro-Fiscal Effects of Climate Shocks From Billions of Geospatial Weather Observations.” IMF Working Papers 2022.156, International Monetary Fund, Washington, DC.
Alker A., 2009. Impact of Global Warming on Infectious Disease in Cyprus. Cyprus Climate Change Assessment Papers – 2009 Compiled by the Master’s Degree Candidates of the Cyprus International Institute in the “Cyprus Climate Workshop”, as part of the Institute’s “Global Climate Update” course, June 2009
Bellon, Matthieu, and Emanuele Massetti, 2022a. “Economic Principles for Integrating Adaptation to Climate Change into Fiscal Policy.” IMF Staff Climate Note 2022/001, International Monetary Fund, Washington, DC.
Bellon, Matthieu, and Emanuele Massetti, 2022b. “Planning and Mainstreaming Adaptation to Climate Change in Fiscal Policy.” IMF Staff Climate Note 2022/003, International Monetary Fund, Washington, DC.
Bosello, Francesco, Robert J. Nicholls, Julie Richards, Roberto Roson, and Richard SJ Tol. “Economic impacts of climate change in Europe: sea-level rise.” Climatic change 112 (2012): 63–81.
Diaz, Delavane B. “Estimating global damages from sea level rise with the Coastal Impact and Adaptation Model (CIAM).” Climatic Change 137, no. 1 (2016): 143–156.
Delforge, D. et al.: EM-DAT: The Emergency Events Database, Preprint, https://doi.org/10.21203/rs.3.rs-3807553/v1, 2023.
Gonguet, Fabien, Claude Wendling, Ozlem Aydin, and Bryn Battersby. 2021. “Climate-Sensitive Management of Public Finances—”Green PFM.” IMF Staff Climate Note 2021/002, International Monetary Fund, Washington, DC.
Harris, Ian, Timothy J. Osborn, Phil Jones, and David Lister, 2020. “Version 4 of the CRU TS monthly high-resolution gridded multivariate climate dataset.” Scientific data 7(1): 109.
Hinkel, Jochen, Jeroen CJH Aerts, Sally Brown, Jose A. Jiménez, Daniel Lincke, Robert J. Nicholls, Paolo Scussolini, Agustín Sanchez-Arcilla, Athanasios Vafeidis, and Kwasi Appeaning Addo. “The ability of societies to adapt to twenty-first-century sea-level rise.” Nature Climate Change 8, no. 7 (2018): 570–578.
Kahn, Matthew E., Kamiar Mohaddes, Ryan NC Ng, M. Hashem Pesaran, Mehdi Raissi, and Jui-Chung Yang. “Long-term macroeconomic effects of climate change: A cross-country analysis.” Energy Economics 104 (2021): 105624.
Kostopoulou, E., C. Giannakopoulos, M. Hatzaki, A. Karali, P. Hadjinicolaou, J. Lelieveld and M.A. Lange (2014) Spatial and temporal patterns of recent and future climat extremes in the Eastern Mediterranean and Middle East region. Nat. Hazards Earth Syst. Sci. 14, 1565–1577
Kopp, Robert E., Frederik J. Simons, Jerry X. Mitrovica, Adam C. Maloof, and Michael Oppenheimer. “Probabilistic assessment of sea level during the last interglacial stage” Nature 462, no. 7275 (2009): 863–867.
Lincke, Daniel, and Jochen Hinkel. “Economically robust protection against 21st century sea-level rise.” Global environmental change 51 (2018): 67–73.
Massetti, Emanuele, and Filippos Tagklis, 2023. “Reference Guide on the FADCP Climate Dataset: Temperature and Precipitation.” Unpublished manuscript, forthcoming, International Monetary Fund, Washington, DC.
Muñoz Sabater, J. (2019): ERA5-Land hourly data from 1950 to present. Copernicus Climate Change Service (C3S) Climate Data Store (CDS). DOI: 10.24381/cds.e2161bac
Ozlem, Aydin, Bryn Battersby, Fabien Gonguet, Claude Wendling, Jacques Charaoui, Murray Petrie, and Suphachol Suphachalasai (2022). How to Make the Management of Public Finances Climate-Sensitive—”Green PFM” Fiscal Affairs Department, How to Note, Washington, DC: International Monetary Fund, 2022.
Slangen, A.B.A., Carson, M., Katsman, C.A. et al. Projecting twenty-first century regional sea-level changes. Climatic Change 124, 317–332 (2014).
Tinbergen, J., 1954, De economische balans van het Deltaplan, bijlage bij het rapport van de Deltacommissie.
Vousdoukas M., Mentaschi L., Mongelli I., Ciscar J-C, Hinkel J.(a), Ward P.(b), Gosling S.(c) and Feyen L., Adapting to rising coastal flood risk in the EU under climate change, EUR 29969 EN, Publications Office of the European Union, Luxembourg, 2020, ISBN 978–92–76–12990–5, doi:10.2760/456870, JRC118512
Vousdoukas, M. I., Mentaschi, L., Voukouvalas, E. et al. (2018). Climatic and socioeconomic controls of future coastal flood risk in Europe. Nature Clim Change 8, 776–780. https://doi.org/10.1038/s41558–018–0260–4
Zittis, G., Almazroui, M., Alpert, P., Ciais, P., Cramer, W., Dahdal, Y., et al. (2022). Climate change and weather extremes in the Eastern Mediterranean and Middle East. Reviews of Geophysics, 60, e2021RG000762. https://doi.org/10.1029/2021RG000762
2022 8th National Communication and 5th Biennial Report under the UNFCCC of Cyprus.
Annex I. The Coastal Impacts and Adaptation Model
1. The Coastal Impact and Adaptation Model (CIAM) is a global optimization model for cost-benefit analysis of adaptation to sea-level rise (SLR) (Diaz, 2016). The model starts from coastal characteristics on more than 12,000 coastal segments covering the entire global coastline from the Dynamic Integrated Vulnerability Assessment model (DIVA), a tool to assess the biophysical and socioeconomic impacts of SLR. The DIVA tool estimates the SLR impacts by considering coastal erosion, coastal flooding, wetland change and saltwater intrusion (DIVA Modeling Framework). For each coastal segment, CIAM develops economic, population, and SLR scenarios. The SLR scenarios are from Kopp et al (2014). On top of SLR, the model considers the expected value of storm surges to include rare but potentially high-impact events.
2. Using data on costs of alternative coastal protections, land values, value of assets along the coastline, the value of statistical life, willingness to pay for preservation of coastal ecosystems, and assumptions on the social cost of relocation. CIAM determines costs and benefits of alternative adaptation strategies (including no adaptation) for each coastal segment. The efficient (optimal) coastal protection strategy has the largest net present value of among all strategies considered by the model. Protection can be full (excluding any inundation also under extreme storm surges), partial (accepting some storm costs) or minimal (with capital and population progressively moving away from the coastline).
3. The model can be used to develop insights on different costs from SLR (Table A1) and different protection strategies (Table A2) in each coastal segment or at higher levels of aggregation.
4. Several impacts have been omitted from the analysis, which could lead to underestimate the cost of no protection, but protection and retreat decisions are fixed for many decades, which could lead to overestimate adaptation costs. Omitted impacts include saltwater intrusion, ocean acidification, erosion, coastal tourism, loss of recreation. The model assumes that protection and retreat decisions are made at the beginning of the planning horizon and remain constant for several decades. A smoother response to SLR would reduce adaptation costs.
5. The general equilibrium effects of all impacts are also omitted, with uncertain effects on cost estimates. Several studies have used global General Equilibrium Models to calculate the macro-economic cost of sea-level rise impacts and adaptation costs (e.g., Bosello et al. 2012). In some countries secondary effects lead to an increase of costs estimated by simulation models like CIAM. In other countries, global market effects (redistribution of tourism flows for example) lead to positive secondary effects.
Cyprus: Cost Specification in the CIAM
Measurement of the cost per period. The model optimizes the adaptation strategies over 50 years.
Cyprus: Cost Specification in the CIAM
| Category | Explanation of the Costs | Measurement of Cost1 |
|---|---|---|
| Protection Cost | Constructing and maintaining protection (generalized as sea walls) to shield the land behind the sea walls from the SLR-caused inundation. | A function of the coastline length, the height of the sea walls, and the value of the land occupied by the sea walls. |
| Retreat Cost | Relocation of population and assets from the affected areas, including the forced emigration in case of no planned adaptation and the planned retreat. | The cost of relocating population and mobile capital in the incremental area of retreat, as well as the cost of demolishing the immobile capital. |
| Inundation Cost | The loss of land and assets due to the SLR-caused inundation. In case of no planned retreat the full value of capital assets is lost. In case of planned retreat only the depreciated level of capital assets is lost. | The cost is based on the extent of land endowment lost and the associated value of the land and the capital stock |
| Wetland Loss Cost | The loss of wetland due to the inability to migrate inland naturally, constrained by the rate of SLR and the lack of space. | The wetland losses take two forms, (1) the total service value of the wetland occupied by the sea walls; (2) the service value of the wetland lost related to the rate of SLR. |
| Storm Cost | The damage to population and assets due to floods caused by storm surge. | The expected damage associated with the risk of the storm surge. The likelihood of storm surge follows the generalized extreme value distribution by using the local surge frequency data from the DIVA tool, while the total land affected by the extreme surge depends on the elevation exposed to a given flood water height. |
Measurement of the cost per period. The model optimizes the adaptation strategies over 50 years.
Cyprus: Cost Specification in the CIAM
| Category | Explanation of the Costs | Measurement of Cost1 |
|---|---|---|
| Protection Cost | Constructing and maintaining protection (generalized as sea walls) to shield the land behind the sea walls from the SLR-caused inundation. | A function of the coastline length, the height of the sea walls, and the value of the land occupied by the sea walls. |
| Retreat Cost | Relocation of population and assets from the affected areas, including the forced emigration in case of no planned adaptation and the planned retreat. | The cost of relocating population and mobile capital in the incremental area of retreat, as well as the cost of demolishing the immobile capital. |
| Inundation Cost | The loss of land and assets due to the SLR-caused inundation. In case of no planned retreat the full value of capital assets is lost. In case of planned retreat only the depreciated level of capital assets is lost. | The cost is based on the extent of land endowment lost and the associated value of the land and the capital stock |
| Wetland Loss Cost | The loss of wetland due to the inability to migrate inland naturally, constrained by the rate of SLR and the lack of space. | The wetland losses take two forms, (1) the total service value of the wetland occupied by the sea walls; (2) the service value of the wetland lost related to the rate of SLR. |
| Storm Cost | The damage to population and assets due to floods caused by storm surge. | The expected damage associated with the risk of the storm surge. The likelihood of storm surge follows the generalized extreme value distribution by using the local surge frequency data from the DIVA tool, while the total land affected by the extreme surge depends on the elevation exposed to a given flood water height. |
Measurement of the cost per period. The model optimizes the adaptation strategies over 50 years.
Cyprus: Adaptation Strategies in the CIAM
Cyprus: Adaptation Strategies in the CIAM
| Category | Explanation |
|---|---|
| No Planned Adaptation | As sea-level rises inundation and floods due to storm surges damage capital and force people to relocate. All asset values in inundated land are lost. Forced retreat is more costly than planned retreat. |
| Protection | A mix of coastal protection barriers which include dikes, floodgates, coastal dunes, etc… Protection is assumed to be 100 percent reliable. Protection can be sized to sea-level only, or to also avoid 10-year, 100-year, 1,000-year, or 10,000-year storm surge levels. Country-level costs in the Optimal Protection scenario presented in Box 1 are calculated by summing costs for the least-cost protection grade at segment level. Protection is planned over a 50-year time horizon in which the social planner has perfect foresight. After this period protection is re-optimized for the following 50 years. |
| Planned Retreat | Coastal land is intentionally abandoned. Population moves and capital is rebuilt behind the retreat perimeter. Immobile capital is efficiently depreciated in preparation of planned retreat. The population incurs relocation costs. The retreat perimeter can be chosen to ensure protection against sea-level rise alone, or to also avoid 10-year, 100-year, 1,000-year, or 10,000-year storm surge levels. |
| Both protection and planned retreat are planned until 2050 and from 2050 to 2100. Over these limited time horizons there is perfect foresight and sea-level is known with certainty. There is no edging against uncertain sea-level rise. | |
Cyprus: Adaptation Strategies in the CIAM
| Category | Explanation |
|---|---|
| No Planned Adaptation | As sea-level rises inundation and floods due to storm surges damage capital and force people to relocate. All asset values in inundated land are lost. Forced retreat is more costly than planned retreat. |
| Protection | A mix of coastal protection barriers which include dikes, floodgates, coastal dunes, etc… Protection is assumed to be 100 percent reliable. Protection can be sized to sea-level only, or to also avoid 10-year, 100-year, 1,000-year, or 10,000-year storm surge levels. Country-level costs in the Optimal Protection scenario presented in Box 1 are calculated by summing costs for the least-cost protection grade at segment level. Protection is planned over a 50-year time horizon in which the social planner has perfect foresight. After this period protection is re-optimized for the following 50 years. |
| Planned Retreat | Coastal land is intentionally abandoned. Population moves and capital is rebuilt behind the retreat perimeter. Immobile capital is efficiently depreciated in preparation of planned retreat. The population incurs relocation costs. The retreat perimeter can be chosen to ensure protection against sea-level rise alone, or to also avoid 10-year, 100-year, 1,000-year, or 10,000-year storm surge levels. |
| Both protection and planned retreat are planned until 2050 and from 2050 to 2100. Over these limited time horizons there is perfect foresight and sea-level is known with certainty. There is no edging against uncertain sea-level rise. | |
Urbanization denotes the growth and expansion of urban areas through population influx, infrastructure development, and changes in land use, resulting in the transformation of rural regions into urban environments.
Extreme heat refers to exceptionally high temperatures that exceed typical seasonal norms.
Extended durations during which an area experiences significantly below-average precipitation, leading to water scarcity, agricultural stress, and environmental impacts over an extended timeframe.
The SSP1–2.6 IPCC scenario keeps global temperature increase above its pre-industrial level below 2°C at the end of the century. In the SSP2–4.5 IPCC scenario emissions continue increasing in line with the continuation of present trends and stabilize at the end of the century. This scenario assumes that climate mitigation policies continue along the observed trend, but countries do not take more aggressive actions to fulfill their Paris commitments. In the SSP3–7.0 IPCC scenario. rather than intensifying climate mitigation efforts, countries start scaling back their implemented policies in a world with limited energy efficiency improvements and continued use of fossil fuels.
This is in line with sea-level rise projections in the 2022 8th National Communication and 5th Biennial report under the UNFCCC of Cyprus, and Table 1 in Deliverable 4.4b “Report on storm surge levels for different scenarios and different time horizon” of the SOCLIMPACT EU-funded project.
Source: CLIMADA, Climate Impact Explorer of Climate Analytics for the NGFS. Cyprus page, consulted on 3/7/2024.
