## wpiea2019178-print-pdf

## Source details

**Canonical URL:** [wpiea2019178-print-pdf](https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019178-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2019/wpiea2019178-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2019/wpiea2019178-print-pdf.pdf.json)

---

### Introduction and background
- Average annual global temperature has increased by about 1°C in the past 40 years.
- In the absence of major cuts to greenhouse gas emissions, it could rise by 4°C or more by 2100 (IPCC, 2014).
- Rising temperatures reduce economic output in countries with hot climates by lowering productivity, investment and labor supply (Burke, Hsiang and Miguel 2015; Acevedo Mejia et al. 2018).
- Low-income countries are overwhelmingly situated in hot regions, contribute little to greenhouse gas emissions, and bear the brunt of the negative economic costs of climate change.
- Adaptation is urgent, particularly for low-income countries.

### Study objective and evidence gap
- Limited evidence on effectiveness of policies attenuating negative effects of excessive heat.
- Paper objective: present empirical evidence on the extent to which macroeconomic and structural policies, institutions and other country characteristics can mitigate the negative relationship between temperature and economic output at the country level.
- Offers a comprehensive cross-country investigation of a wide range of potentially useful policies.

### Key empirical findings (summary)
- Short-run policy buffers that help:
  - Low public debt to GDP
  - Foreign aid
  - Remittances
- Long-run structural and institutional characteristics that help:
  - Exchange rate flexibility
  - High financial sector liberalization
  - Good infrastructure
  - Democratic institutions
  - Low inequality
- Subnational evidence:
  - Hot regions in high-income countries on average sustain less economic damage than hot regions in low-income countries.
- Overall finding:
  - No evidence of fully successful adaptation at the macroeconomic level in the past 40 years.

### Climate adaptation policy toolkit (functions and instruments)
- Three broad functions of policies:
  - Enhance ability to smooth the impact of shocks (policy buffers, social safety nets, exchange rate flexibility, financial sector policies).
  - Enhance flexibility and foster structural transformation (labor market policies, education and health, infrastructure, institutions).
  - Reduce exposure and vulnerability through soft and hard adaptation measures (public information, early warning systems, building codes, fiscal incentives, climate-smart infrastructure, private and sovereign insurance, multilateral risk-sharing).
- Examples of policy instruments:
  - Fiscal buffers: lower public debt as a share of GDP.
  - Monetary buffers: low inflation.
  - High international reserves.
  - High foreign aid inflows.
  - High remittances.
  - Exchange rate flexibility (de facto regime not pegged).
  - Labor market policies, education and health, financial sector policies, infrastructure, institutional reforms.
  - Soft measures: public information, early warning systems, zoning, fiscal incentives.
  - Hard measures: retrofitting, irrigation, drainage, seawalls, climate-smart infrastructure.
- Note: Appropriate adaptation measures are highly specific to local climate risks; lack of comparable data precludes cross-country empirical analysis of specific adaptation measures.

### Empirical approach and data
- Key data sources: IMF World Economic Outlook (WEO); World Bank World Development Indicators (WDI); University of East Anglia’s Climate Research Unit (CRU) historical temperature and precipitation databases.
- Per capita GDP: WEO and WDI.
- Historical temperature and precipitation: CRU.
- Temperature and precipitation aggregation: CRU grid-cell data at 0.5 × 0.5 degree resolution aggregated to country level using 1950 population in each cell as weights.
- Baseline method: Jordà’s (2005) local projection to trace impulse response of real per capita GDP to a weather shock; cumulative growth estimated for horizons h = 0 to 7.
- Controls: country fixed effects and region-year fixed effects; standard errors clustered at the country level.
- Weather enters quadratically to capture non-linearities; marginal effect of a 1°C increase: ∂(y_{i,t+h} − y_{i,t−1})/∂T_{i,t} = β1_h + 2 β2_h T_{i,t}.
- Table 1 (h=0, contemporaneous) reported entries:
  - Temperature 1.347 *** (0.357); Temperature2 –0.051 *** (0.011).
  - Precipitation 0.110 (0.104); Precipitation2 –0.003 (0.002).
  - Panel B impacts at median temperatures:
    - AE (T=11°C): 0.218 (0.196)
    - EM (T=22°C): –0.911 *** (0.264)
    - LIDC (T=25°C): –1.219 ***
  - Threshold Temperature (°C) 13°
  - Adjusted R2 0.14 (column 1) and 0.09 (column 2)
  - Number of Countries 189 (column 1) and 127 (column 2)
  - Number of Observations 8,815 (column 1) and 6,135 (column 2)
- Reported threshold range where temperature effect switches from positive to negative: between 13°C – 15°C.
- Baseline finding: warming boosts output in cooler countries and suppresses growth in hotter countries; low-income countries (warmer climates) face much more deleterious effects that remain significant up to seven years after the shock.

### Role of domestic policies and institutions — empirical strategy
- Two approaches:
  1. Include interactions between weather shock and policy variable in projection regressions.
  2. Restrict sample to countries with average annual temperature exceeding 15°C.
- Policy variables are mostly lagged and generally transformed into indicator variables (above/below median) for interpretation.
- Measurement of buffers (exact criteria):
  - Fiscal buffers: public debt as a share of GDP is less than the 75th percentile.
  - Monetary buffers: annual inflation is less than 10 percent.
  - High international reserves: reserves minus gold can cover at least four months of imports.
  - High foreign aid: foreign aid inflows as a share of GDP are in the 75th percentile.
  - High remittances: per capita remittances greater than the 75th percentile.
- Exchange rate policy indicator: de facto exchange rate regime is not pegged (Reinhart and Rogoff (2004) classification).
- Caveat: interaction coefficients are difficult to interpret causally; policy variation is not random and may correlate with omitted time-varying attributes.
- Presentation: Figures 3 and 4, Tables 2 and 3 report estimated effect of a 1°C increase on per capita output at horizons 0–7 under policy/no-policy scenarios; p-values test differences. Gray areas in figures indicate horizons where lines differ at the 15 percent level. Statistical significance markers: * p<0.1; ** p<0.05; *** p<0.01.

### Empirical moderation findings (patterns)
- Sample: countries with average temperature > 15°C.
- Short-run buffers associated with attenuation (often short-lived and not always statistically significant):
  - Low public debt, high foreign aid, high remittances show mitigation in some horizons.
  - Countries with non-pegged exchange rates tend to recover faster from weather shocks.
- Structural policies and institutions associated with medium-term attenuation (standard errors often large; outer-horizon point estimates larger):
  - Better regulated domestic and international financial markets.
  - Wider infrastructure availability.
  - Strong democratic institutions.
  - Lower income inequality.
- Figures and tables report horizon-by-horizon marginal effects; specific quantitative moderation magnitudes and p-values are horizon-specific and reported in Tables 2 and 3 (not reprinted here).

### Inequality, short-term effects and policy buffers
- Short-term negative effects of a 1°C temperature shock tend to be larger in countries with lower buffers; larger estimated responses appear in columns (2), (5), and (8) of Table 2.
- Differences in short-term responses by buffers are typically not statistically significant; where they are (fiscal buffers, foreign aid, remittances) the effects tend to be very short-lived.
- Exchange rate flexibility associated with better adjustment after shocks (e.g., windstorms and earthquakes).
- Identification: estimates on countries with average annual temperature above 15°C; interactions with temperature, precipitation and their lags; country and region-year fixed effects; standard errors clustered at the country level.

### Role of development — subnational evidence
- Subnational growth data: roughly 1,460 provinces and states across 79 countries (Gennaioli et al. 2014).
- Focus sample: provinces/states with average temperature > 15°C: about 610 provinces and states (about 607 reported).
- Regression: indicator p_{i,t} = 1 for states/provinces in advanced economies interacted with lag of growth; standard errors clustered at the province level.
- Key finding: Temperature shocks hurt hot areas in emerging market and developing economies significantly more than in advanced economies (Figure 5; Table 4).
- Development appears to provide some insulation from temperature shocks.

### Historical adaptation and readiness
- Rolling 20-year estimations for a median low-income developing country (average annual temperature 25 degrees Celsius) show contemporaneous response of per capita output to temperature shocks has remained essentially constant over time.
- Interpretation: over the past 60 years, median low-income countries show little evidence of overall adaptation that mutes GDP response to temperature shocks.
- Constraints to adaptation in low-income countries: high costs, limited access to credit, insufficient information on benefits, limited planning rationality, inadequate access to technology.
- Notre Dame Global Adaptation Index measures readiness and capacity based on 45 indicators.
- Using ND-GAIN, low-income countries—those with greatest need—are the least prepared to adapt to weather shocks.

### DIG model complement — purpose and structure
- Model: Debt, Investment and Growth (DIG) model (Buffie et al. 2012); optimizing intertemporal small open economy model with perfect foresight; traded and non-traded sectors; public capital productive in both sectors.
- Features:
  - Cobb-Douglas production; firms combine labor, private capital, and public capital.
  - Households: savers and hand-to-mouth consumers in fixed proportions; only savers maximize lifetime utility.
  - Government: spending on transfers, debt service, and partially inefficient infrastructure investment; revenue from VAT and user fees; deficit financed through domestic and external borrowing (concessional and commercial); accepts all concessional loans offered; no sovereign default allowed; taxes and transfers adjust.
  - Shocks: weather shocks affect total factor productivity (TFP) and investment in productive capital; TFP evolution is exogenous.
- Calibration goal: shape and maximum decline of output in model broadly follow estimated response of GDP per capita to a 1°C temperature shock in a representative low-income country (baseline temperature of 25°C).

### Parameterization (Table 5 reproduced)
- Parameter — Value (percent)
  - Initial Return on Infrastructure Investment — 30
  - Public Domestic Debt-to-GDP Ratio — 10
  - Public Concessional Debt-to-GDP Ratio — 30
  - Public External Commercial Debt-to-GDP Ratio — 5
  - Oil Revenues-to-GDP Ratio — 2
  - Real Interest Rate on Public Domestic Debt — 7
  - Real Interest Rate on Public External Commercial Debt — 4
  - Trend per Capita Growth Rate — 2.8

### Model simulation findings — policies and mechanisms
- Fiscal space (proxied by transfers from advanced economies):
  - Additional transfers of 1 percent of the recipient country’s GDP reduce the depth of the recession by about 0.5 percent throughout the simulation period.
  - Transfers increase public infrastructure stock and boost productive capacity in both traded and non-traded sectors, raising output short and long term.
  - Baseline concessional debt equal to 30 percent of GDP; one-time 1 percent of GDP transfers are not overwhelmingly large but significant short-run.
- Public investment efficiency:
  - Efficiency estimates cited: low efficiency of 20 percent to average efficiency of 60 percent.
  - With high public investment efficiency, additional transfers of 1 percent of GDP can effectively dampen adverse consequences of a weather shock.
  - With low public investment efficiency, additional transfers make little difference.
- Cost and speed of factor reallocation:
  - Higher capital reallocation costs slow recovery; if climate shock mostly destroys private capital rather than lowering TFP, recovery is slower and GDP damage larger.
  - Capital adjustment cost captured by elasticity of investment with respect to Tobin’s q (higher elasticity implies lower adjustment costs).
- Labor reallocation, unemployment, and hysteresis:
  - Hysteresis captured via sensitivity of productivity to lagged negative output gaps; calibration uses elasticity of current wages to lagged hours worked of 0.2.
  - Hysteresis can significantly prolong and deepen effects of weather shocks.
  - Policy implication: preserve human capital via incentives for unemployed to participate in human-capital-preserving activities (example: public works such as Ethiopian Productive Safety Net Program).

### Investment in adaptation strategies — model extension
- Adaptation investments are often public goods; government involvement required due to nonrival and nonexcludable payoffs.
- Data limitations: poor cross-country data on adaptation investment; empirical analysis found no aggregate evidence of successful adaptation.
- Model extension (Annex 2): government provides fiscal incentives (subsidies) when private adaptation expenditure falls 20 percent short of social optimum; subsidies restore optimality.
- Quantitative simulation result:
  - Over 20 years, each $1 spent on adaptation subsidies by the government reduces aggregate weather damage by $2.
  - Mechanism: reduced weather-related productivity losses spur private investment response, boosting GDP in medium and long term.

### Annex modeling results (optimal adaptation theoretical framework)
- Damage and adaptation functions (as specified):
  - Gross damage f jt jt jt GD gdT q ==
  - Firm i capacity to adapt increases in firm protection expenditures ,i jt AD and total sectoral protection expenditures.
  - Residual damage: , , ( ,) jt i jt jt i jti jt gd OADAD φ Ω =
  - Marginal damage reduction from adaptation spending is decreasing; φ is elasticity of damage reduction to adaptation.
  - Firm-level optimal adaptation (symmetric equilibrium):
    - ( ) 1 11 , , . jt i jt AD t GD AD P φς φ ++  =   
  - Firm-level residual damage: ( ) 1 , jt jt jt gd AD φς+ Ω =
  - Social planner optimal adaptation:
    - ( ) ( ) 1 11 , 1. jt SP jt AD t GD AD P φς φς ++  = +   
  - Adaptation spending gap expressed as a fraction of the socially optimal adaptation spending (formula reproduced in source).
  - Government subsidy per unit cost required to achieve social optimum: ,jtς υ with () , 1 jtς ς υ ς = +

### Conclusions and policy priorities
- Empirical and model evidence: policies help mitigate negative effects of weather shocks on output, but effects are typically marginal rather than fully offsetting.
- Empirical correlates of reduced weather-related output losses:
  - Fiscal buffers: low public debt, high foreign aid, high remittances.
  - Structural policies and institutions: high financial sector liberalization, low capital account restrictions, good infrastructure, high polity score.
- Alarm: magnitudes of estimated effects of adaptation policies are small; none of the specific policies identified are sufficient to completely erase negative effect of temperature increases on aggregate output.
- Exception: subnational analysis suggests higher overall development associated with smaller negative effect of temperature increases, implying advanced countries’ “hot” regions are better insulated.
- Policy recommendations and priorities:
  - Strengthen fiscal buffers (low public debt, build fiscal space).
  - Improve public investment efficiency and public sector governance to ensure transfers translate into productive infrastructure.
  - Facilitate factor reallocation (financial liberalization, reduce bureaucratic impediments) to speed recovery.
  - Preserve human capital (public works and human-capital-preserving programs) to limit hysteresis.
  - Public support for private adaptation investment (subsidies) can be cost-effective (each $1 can yield $2 aggregate damage reduction over 20 years in model).
  - Greater international commitment to climate change mitigation is needed to limit future temperature increases; supporting low-income countries in adaptation is both moral duty and sound global economic policy.

*Source: wpiea2019178-print-pdf (IMF staff; Notre Dame Global Adaptation Index).*

### INTRODUCTION

### INTRODUCTION

### Background and urgency
- The average annual global temperature has increased by about 1°C in the past 40 years.
- In the absence of major cuts to greenhouse gas emissions, it could rise by 4°C or more by 2100 (IPCC, 2014).
- Rising temperatures reduce economic output in countries with hot climates by lowering productivity, investment and labor supply (Burke, Hsiang and Miguel 2015; Acevedo Mejia et al. 2018).
- Low-income countries are overwhelmingly situated in hot regions, contribute little to greenhouse gas emissions, and bear the brunt of the negative economic costs of climate change.
- Adaptation is urgent, particularly for low-income countries.

### Evidence gap and study objective
- Policymakers and researchers have limited evidence about the effectiveness of policies in attenuating the negative effects of excessive heat.
- The paper aims to fill this gap by presenting empirical evidence on the extent to which macroeconomic and structural policies, institutions and other country characteristics can mitigate the negative relationship between temperature and economic output at the country level.
- The analysis offers a comprehensive and unified cross-country investigation of a wide range of potentially useful policies.

### Key empirical findings
- Short-run policy buffers that help:
  - Low public debt to GDP
  - Foreign aid
  - Remittances
- Long-run structural and institutional characteristics that help:
  - Exchange rate flexibility
  - High financial sector liberalization
  - Good infrastructure
  - Democratic institutions
  - Low inequality
- Using subnational data:
  - Hot regions in high-income countries on average sustain less economic damage than hot regions in low-income countries, suggesting general economic development policies would complement climate adaptation strategies.
- Overall finding:
  - No evidence of fully successful adaptation at the macroeconomic level in the past 40 years.

### Model complement and simulation results
- The empirical analysis is complemented with a dynamic general equilibrium model based on the Debt, Investment and Growth (DIG) model of Buffie et al. (2012).
- The DIG model:
  - Captures characteristics pertinent to low-income countries—such as low public investment efficiency and high capital adjustment costs—and incorporates the structural transformation process.
  - Is preferable for studying the impact of climate change in low-income countries relative to the Integrated Assessment Models (IAMs) more commonly used to assess climate change effects.
- Model purpose and results:
  - Used to simulate how specific macroeconomic policies and structural transformation help reduce the negative effect of temperature shocks.
  - Addresses potential endogeneity concern that only countries with the most severe climate shocks introduce adaptation policies, which could bias empirical assessments of policy effectiveness.
  - Simulations are consistent with empirical findings about which specific policies might help.
  - Simulations confirm that none of the policies can fully insulate countries from the negative effects of heat shocks.

### Literature context and nuanced findings
- Prior literature on adaptation to weather shocks shows mixed evidence:
  - Positive/mitigating roles identified for:
    - Greater financial development
    - Greater insurance penetration
    - Higher quality institutions and democracy
    - Flexible exchange rates
    - Overall development
    - Climate-smart technologies (e.g., cyclone risk adaptation) (Hsiang and Narita 2012; Hsiang and Jina 2014)
  - Studies finding little or no adaptation for repeated negative weather shocks, especially at aggregate output or growth levels:
    - The negative response of GDP to temperature shocks did not diminish over time (Burke, Hsiang, Miguel 2015; Dell, Jones, Olken 2012).
    - Lack of adaptation in agriculture (Schlenker and Roberts 2009; Burke and Emerick 2016), labor productivity (Heal and Park 2013), conflict (Burke et al. 2009), and crime (Ranson 2014).
  - Potential for future adaptation:
    - Comparative advantages and international trade adjustments could limit negative effects if countries adapt crop production accordingly (Costinot et al. 2016).
- Micro versus macro adaptation trade-offs:
  - Individual-level adaptations such as greater air-conditioning uptake have been successful in reducing heat exposure (Deschenes and Greenstone 2011; Barreca et al. 2016).
  - Scaling such adaptations can produce negative externalities in emerging markets reliant on fossil-fuel electricity—worsening pollution and future climate risks (Davis and Gertler 2015)—and local pockets of hot air from exhaust, affecting densely populated urban areas.

### Policy implications and overall conclusion
- Policies and structural reforms can play a role in reducing the negative economic effects of higher temperatures, both in the short run and the long run.
- However, the analysis finds that policies studied are not sufficient to eliminate the negative effect of temperature increases.
- Ultimate, permanent relief from future negative climate shocks requires limiting greenhouse gas emissions, which the scientific community agrees would limit further global warming.

### Paper organization (as stated)
- Section II describes the different types of policies that are studied.
- Section III introduces the data and outlines the empirical approach and findings.

*Source: wpiea2019178-print-pdf - INTRODUCTION*

### Section IV discusses our model approach and findings; and Section V concludes. Annexes

### CLIMATE ADAPTATION POLICY: A TOOLKIT

### Toolkit overview
- The toolkit includes domestic policy actions and private choices to cope with weather shocks and climate change risks. These range from macroeconomic and structural policies (e.g., fiscal buffers, social safety nets) to country-specific adaptation strategies (e.g., climate-smart infrastructure investments) and sovereign insurance (e.g., catastrophe bonds). Private migration is noted as an extreme form of adaptation.
- Policies serve three broad functions:
  - Enhance ability to smooth the impact of shocks (policy buffers, well-targeted social safety nets, exchange rate flexibility, financial sector policies).
  - Enhance flexibility and foster structural transformation (labor market policies, education and health policies, infrastructure investment, strong institutional framework).
  - Reduce exposure and vulnerability through soft and hard adaptation measures (public information provision, early warning systems, building codes and land use planning, fiscal incentives for adaptive technologies, climate-smart infrastructure, private and sovereign insurance, multilateral risk-sharing mechanisms).

### Examples and policy instruments (as summarized)
- Policy buffers and short-run smoothing:
  - Fiscal buffers (e.g., lower public debt as a share of GDP).
  - Monetary buffers (e.g., low inflation).
  - High international reserves.
  - High foreign aid inflows.
  - High remittances.
  - Exchange rate flexibility (de facto regime not pegged).
- Structural and long-run adaptation enablers:
  - Labor market policies to facilitate movement across sectors/regions.
  - Education and health to strengthen human capital, reduce mortality/morbidity from heat, and facilitate lifelong learning.
  - Financial sector policies to ensure access to credit, insurance, and other financial services.
  - Infrastructure investment and strong institutional frameworks.
- Specific adaptation measures:
  - Soft measures: public information provision about climate-related risks; early warning systems and evacuation schemes; stronger building laws, land use planning, zoning; fiscal incentives and appropriate pricing for adaptive technologies.
  - Hard measures: retrofitting properties; irrigation, drainage; seawalls; climate-smart infrastructure.
- Note: Appropriate adaptation measures are highly specific to local climate risks; lack of comparable data precludes cross-country empirical analysis of specific adaptation measures.

### Empirical analysis: data and aggregation
- Key data sources: IMF World Economic Outlook (WEO); World Bank World Development Indicators (WDI); University of East Anglia’s Climate Research Unit (CRU) historical temperature and precipitation databases.
- Per capita GDP: WEO and WDI.
- Historical temperature and precipitation: CRU.
- Temperature and precipitation aggregation: CRU grid-cell data at 0.5 × 0.5 degree resolution aggregated to country level using 1950 population in each cell as weights (captures the average weather experienced by a person in the country).

### Baseline specification (local projection approach)
- Method: Jordà’s (2005) local projection to trace impulse response of real per capita GDP to a weather shock; uses within-country and across-country year-to-year fluctuations in temperature and precipitation.
- Regressions estimate cumulative growth of real GDP per capita between horizons t − 1 and t + h, with h from 0 (contemporaneous) to 7 (7 years after the shock). Country fixed effects and region-year fixed effects control for time-invariant differences and common regional shocks. Standard errors clustered at the country level.
- Weather variables enter quadratically to capture non-linearities: in cooler climates increases in temperature may be beneficial; in already-hot climates increases are detrimental.
- Marginal effect of a 1°C increase in temperature at horizon h: ∂(y_{i,t+h} − y_{i,t−1})/∂T_{i,t} = β1_h + 2 β2_h T_{i,t}.
- Table 1 (horizon h=0, contemporaneous) highlights:
  - Temperature coefficients (column 1): Temperature 1.347 *** (0.357); Temperature2 –0.051 *** (0.011).
  - Precipitation coefficients (column 1): Precipitation 0.110 (0.104); Precipitation2 –0.003 (0.002).
  - Panel B: Impact of a 1°C increase in temperature on real output per capita at median temperatures:
    - AE (T=11°C): 0.218 (0.196)
    - EM (T=22°C): –0.911 *** (0.264)
    - LIDC (T=25°C): –1.219 ***
  - Reported table entries (as presented): Threshold Temperature (°C) 13°; Adjusted R2 0.14 (column 1) and 0.09 (column 2); Number of Countries 189 (column 1) and 127 (column 2); Number of Observations 8,815 (column 1) and 6,135 (column 2).
  - Based on the quadratic specification and robustness checks, the threshold temperature where temperature effect switches from positive to negative is reported to range between 13°C – 15°C.
- Key baseline finding:
  - Non-linear relationship confirmed: warming boosts output in cooler countries and suppresses growth in hotter countries. Low-income countries (with warmer climates) face much more deleterious effects; for low-income countries the negative effect remains negative and significant up to seven years after the shock.
- Figures and impulse-response plots:
  - Figure 2 (described): effect on cumulative per capita output from a 1°C increase in temperature estimated at median temperatures for advanced, emerging market, and low-income countries at horizons 0 through 7.

### Role of domestic policies and institutions: empirical strategy
- Two modifications to baseline to study mitigation by policies:
  1. Include interactions between weather shock and policy variable in the projection regressions (equation (3)).
  2. Restrict sample to countries with average annual temperature exceeding 15°C (where temperature increases have a statistically significant linear negative impact).
- Policy variables (p_{i,t}) are mostly lagged and generally transformed into indicator variables (above/below median in estimation sample) for ease of interpretation.
- Measurement of buffers (exact criteria used):
  - Fiscal buffers: public debt as a share of GDP is less than the 75th percentile.
  - Monetary buffers: annual inflation is less than 10 percent.
  - High international reserves: international reserves minus gold can cover at least four months of imports.
  - High foreign aid: foreign aid inflows as a share of GDP are in the 75th percentile.
  - High remittances: per capita remittances in real US dollars received are greater than the 75th percentile.
- Exchange rate policy indicator: de facto exchange rate regime is not pegged (coarse classification of Reinhart and Rogoff (2004)).
- Caveats: coefficients on interaction terms are difficult to interpret causally because policy/institution variation is not random, policies may correlate with omitted time-varying country attributes, and policy data availability varies across countries and time.
- Presentation of results:
  - Figures 3 and 4, and Tables 2 and 3 report estimated effect of a 1°C increase in temperature on per capita output at horizons 0 through 7 under scenarios where the policy is not in place versus in place, and p-values testing differences between scenarios.
  - Figure 3 focuses on role of policy buffers (panels include Public Debt, Inflation, International Reserves, Foreign Aid, Remittances, Exchange Rate Flexibility). Gray areas indicate horizons where the two lines are significantly different at the 15 percent level.
  - Figure 4 focuses on structural policies and institutions (panels include Domestic Financial Sector Reform, International Finance, Human Capital, Physical Capital, Political Regime).

### Key empirical findings on policy moderation (summary of patterns)
- For countries with average temperature > 15°C (sample for interaction analysis), empirical results examine whether the presence of buffers or stronger structural characteristics attenuates negative impacts of a 1°C temperature increase across horizons 0–7.
- The analysis reports horizon-by-horizon marginal effects under alternative policy scenarios and statistical tests of differences; graphical evidence (Figures 3 and 4) highlights instances where policy buffers or structural characteristics significantly alter the temperature–growth relationship (significance indicated at up to the 15 percent level in figures).
- Specific quantitative moderation magnitudes and p-values are reported in Tables 2 and 3 (not reprinted here), and are horizon-specific.

### Research design notes and robustness
- Identification relies on year-to-year within-country and cross-country temperature and precipitation fluctuations, quadratic specification in weather variables, country and region-year fixed effects, and clustered standard errors.
- Robustness checks and extensions are described in Acevedo Mejia et al. (2018) (baseline medium-term effects, channels of impact, and alternative specifications).

*Source: IMF staff compilation from the document content provided.*

### 6. Inequalit  y

### 6. Inequality

### A. Short-term Effects and Policy Buffers
- The short-term negative effects of a 1°C temperature shock tend to be larger in countries with lower buffers; larger estimated responses appear in columns (2), (5), and (8) of Table 2.
- Differences in short-term responses by buffers are typically not statistically significant; where they are (fiscal buffers, foreign aid, and remittances) the effects tend to be very short-lived.
- Exchange rate regime: countries with non-pegged exchange rates tend to recover faster from weather shocks; exchange rate flexibility is associated with better adjustment after windstorms and earthquakes (Ramcharan 2009).
- Empirical setup note: results in Table 2 are estimated on a sample of countries with average annual temperature above 15°C. Indicators for policy measures are interacted with temperature, precipitation, and their lags, controlling for country and region-year fixed effects, lags of growth and policy measure, and forwards of temperature and precipitation. Standard errors are clustered at the country level.
- Statistical significance markers used: * p<0.1; ** p<0.05; *** p<0.01.

### B. Structural Policies and Institutions
- Point estimates indicate that the medium-term adverse effect of a temperature increase fades when:
  - domestic and international financial markets are better regulated,
  - infrastructure is widely available,
  - democratic institutions are strong,
  - the distribution of income is fairly even.
- Standard errors are often large; it is frequently difficult to reject the null that policies have no effect, yet outer-horizon point estimates are substantially larger in columns (2), (5), and (8) of Table 3, suggesting attenuation of damage with stronger structural policies.
- Other adaptation strategies mentioned include migration and financial instruments (private and sovereign insurance, e.g., crop insurance, catastrophe bonds).
- Empirical setup note: Table 3 estimates equation (3) on countries with average annual temperature above 15°C, interacting policy indicators with temperature, precipitation, and their lags; region-year fixed effects and province/country fixed effects included; standard errors clustered at the country level.

### C. The Role of Development (Subnational Evidence)
- Motivation: within-country geographic heterogeneity in advanced economies (e.g., U.S. states from about 7°C to 21°C) allows comparison of hot subnational areas in advanced vs. non-advanced economies.
- Data and sample:
  - Subnational growth data from roughly 1,460 provinces and states across 79 countries (Gennaioli et al. 2014).
  - Focus on provinces/states with average temperature greater than 15°C: about 610 provinces and states (about 607 appears in reporting).
  - Regressions estimate equation (3) with an indicator p_{i,t} = 1 for states/provinces in advanced economies; p_{i,t} is interacted with lag of growth; standard errors clustered at the province level.
- Key findings:
  - Temperature shocks hurt hot areas in emerging market and developing economies significantly more than in advanced economies (Figure 5; Table 4).
  - Development appears to provide some insulation from temperature shocks.
- Table/figure reporting details:
  - Table 4 reports the impact of a 1°C increase in temperature on per capita output for the full sample and separately for Advanced Economies and Non-Advanced Economies.
  - Statistical significance markers: * p<0.1; ** p<0.05; *** p<0.01.

### D. Historical Adaptation and Readiness for the Future
- Historical adaptation evidence:
  - Using rolling 20-year estimations of equation (1) and evaluating effects via equation (2) for a median low-income developing country with average annual temperature of 25 degrees Celsius, the contemporaneous response of per capita output to temperature shocks has remained essentially constant over time.
  - Interpretation: over the past 60 years, median low-income countries show little evidence of overall adaptation that mutes the GDP response to temperature shocks (Figure 6).
  - The contemporaneous response of GDP per capita to weather shocks also has not changed much over time for advanced economies or emerging markets (footnote).
- Constraints to adaptation in low-income countries include: high costs, limited access to credit for financing adaptation, insufficient information on benefits, limited rationality in planning for future risks, and inadequate access to technology (Carleton and Hsiang 2016).
- Readiness metric:
  - The Notre Dame Global Adaptation Index measures readiness and capacity to adapt based on 45 indicators (food, water, health services, ecosystem services, infrastructure, etc.).
  - Using this metric, low-income countries—those with the greatest need—are the least prepared to adapt to weather shocks (Figure 7).
- Policy implication: findings underscore the need for global action to support low-income countries’ adaptation and global efforts to reduce greenhouse gas emissions.

### E. Model-Based Analysis (DIG Model)
- Purpose: Complement empirical analysis by using the Debt, Investment and Growth (DIG) model (Buffie et al. 2012) to illustrate how policies can moderate consequences of temperature increases in low-income countries.
- Model features:
  - Dynamic general equilibrium model of a small open economy with an extensive fiscal component.
  - Agents optimize lifetime utility subject to intertemporal budget constraints with perfect foresight.
  - Two-sector production side using public and private capital as inputs.
  - Model calibrated to features relevant to low-income countries (e.g., low public investment efficiency).
- Advantage: model-based approach permits isolation of specific policy roles and tracing of conceptual channels that are difficult to identify in cross-country empirical work.

*Source: IMF staff calculations.*

### 1. Adaptation Capacity Ind ex (x-axis)

### 1. Adaptation Capacity Ind ex (x-axis)

### Figures and data depiction
- The figure(s) depict the estimated effect of a 1°C increase in temperature on per capita output at horizon 0 against countries’ score for:
  - adaptation readiness (Adaptation Readiness Index on x-axis in Figure 2),
  - adaptation capacity (Adaptation Capacity Index on x-axis in Figure 1).
- Higher score indicates better adaptation capacity and more readiness.
- Country groups shown: Low-income deve lopingcountries; Emergingmarke t economies; Ad va nced economies.
- Axis tick values reproduced exactly as shown:
  - Left/right vertical axis (presumed effect scale): –2.0, –1.5, –1.0, –0.5, 0.0, 0.5, 1.0, 1.5, 2.0.
  - Horizontal tick values in one figure: 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0.

### Sources and note
- Sources: Notre Dame Global Adaptation Index; and IMF staff calculations.
- Note: The figure depicts the estimated effect of a 1°C increase in temperature on per capita output at horizon 0 against countries’ score for adaptation readiness and adaptation capacity. A higher score indicates better adaptation capacity and more readiness.

### Key visual message
- The plotted relationship shows cross-country variation in the immediate (horizon 0) per-capita output response to a 1°C temperature increase, conditional on measured adaptation readiness and capacity scores.

### Final source attribution
*Source: wpiea2019178-print-pdf (IMF staff; Notre Dame Global Adaptation Index).*

### A. Model description and calibration (DIG model)
- Model type: Optimizing intertemporal small open economy model with perfect foresight (DIG model) with traded and non-traded sectors; public capital is productive in both sectors.
- Production: Cobb-Douglas technologies; firms combine labor, private capital, and public capital.
- Households: Savers and hand-to-mouth consumers in fixed proportions; only savers maximize lifetime utility.
- Government: Spending on transfers, debt service, and (partially inefficient) infrastructure investment; revenue from VAT and user fees; deficit financed through domestic and external borrowing (concessional and commercial); accepts all concessional loans offered by official creditors; no sovereign default allowed (taxes and transfers eventually adjust).
- Shocks: Weather shocks affect total factor productivity (TFP) and investment in productive capital; TFP evolution is exogenous.
- Calibration goal: Shape and maximum decline of output in model broadly follow estimated response of GDP per capita to a 1°C temperature shock in a representative low-income country (baseline temperature of 25°C).

### Parameterization (Table 5 reproduced exactly)
- Parameter — Value (percent)
  - Initial Return on Infrastructure Investment — 30
  - Public Domestic Debt-to-GDP Ratio — 10
  - Public Concessional Debt-to-GDP Ratio — 30
  - Public External Commercial Debt-to-GDP Ratio — 5
  - Oil Revenues-to-GDP Ratio — 2
  - Real Interest Rate on Public Domestic Debt — 7
  - Real Interest Rate on Public External Commercial Debt — 4
  - Trend per Capita Growth Rate — 2.8
- Sources for parameters: Buffie et al. (2012); and authors' calculations.
- Majority of parameters follow Buffie et al. (2012) with exceptions to reflect decline in global interest rates, projection of trend GDP growth in low-income countries, and sample median of public-debt-to-GDP ratios.

### B. Role of domestic policies and institutions — model simulations and quantitative findings
- Fiscal space (proxied by transfers from advanced economies):
  - Additional transfers of 1 percent of the recipient country’s GDP reduce the depth of the recession by about 0.5 percent throughout the simulation period.
  - Transfers increase the stock of public infrastructure and boost productive capacity in both traded and non-traded sectors, raising output in both short and long term.
  - Baseline concessional debt equal to 30 percent of GDP; these one-time increases in transfers are characterized as not overwhelmingly large but significant in the short run.
- Public investment efficiency:
  - Efficiency estimates cited: low efficiency of 20 percent to average efficiency of 60 percent (references: Foster and Briceno-Garmendia 2010; Hulten 1996; Pritchett 2000).
  - In simulations, with high public investment efficiency, receipt of additional transfers of 1 percent of GDP can effectively dampen adverse consequences of a weather shock.
  - With low public investment efficiency, receiving additional transfers makes little difference because funds are not transformed into productive infrastructure.
- Cost and speed of factor reallocation:
  - Higher costs of capital reallocation slow recovery from weather shocks.
  - Quantitative impact of adjustment costs of typical magnitude for a low-income country appears small, but results are qualitative guides only.
  - If the climate shock mostly destroys private capital rather than lowering TFP, recovery is slower and damage to GDP larger (rebuilding capital slower than TFP rebound).
  - Ease of factor reallocation captured in model by cost of private capital adjustment parameter: cost of capital adjustment is inversely proportional to elasticity of investment with respect to Tobin’s q (higher elasticity implies lower capital adjustment costs).
- Labor reallocation, unemployment, and hysteresis:
  - Hysteresis (permanent “scarring”) captured via sensitivity of productivity to lagged negative output gaps.
  - Calibration note: elasticity of current wages to lagged hours worked from Altuǧ and Miller (1998) is 0.2, representing a high degree of hysteresis in model specification.
  - Simulations suggest hysteresis could significantly prolong and deepen effects of weather shocks.
  - Policy implication: preserve human capital via incentives for unemployed to participate in human-capital-preserving activities (example: public works such as Ethiopian Productive Safety Net Program).

### C. Investment in adaptation strategies — model extension and results
- Nature of adaptation investments: many are public goods (early-warning systems, information campaigns, green infrastructure) with nonrival and nonexcludable payoffs; government involvement often required due to inability of private agents to internalize full social benefits.
- Data limitations: poor cross-country data on adaptation investment; empirical analysis found no evidence of successful adaptation in the aggregate (response of output per capita to weather shocks remained similar over time).
- Model extension (Annex 2): government provides fiscal incentives (subsidies) when private adaptation expenditure falls 20 percent short of the social optimum; subsidies restore optimality.
- Quantitative simulation result:
  - Over 20 years, each $1 spent on adaptation subsidies by the government reduces aggregate weather damage by $2.
  - Mechanism: reduced weather-related productivity losses spur private investment response, boosting GDP in medium and long term.
- General principle: improving resilience through public support of private investment in adaptation can reduce weather-driven downturns and accelerate recoveries.

### Conclusion — synthesized findings and policy implications
- Empirical and model evidence: policies help mitigate negative effects of weather shocks on output, but effects are typically marginal rather than fully offsetting.
- Empirical correlates of reduced weather-related output losses:
  - Fiscal buffers: low public debt, high foreign aid, high remittances.
  - Structural policies and institutions: high financial sector liberalization, low capital account restrictions, good infrastructure, high polity score.
- Alarm: magnitudes of estimated effects of adaptation policies are small; none of the specific policies identified are sufficient to completely erase negative effect of a temperature increase on output at the country-aggregate level.
- Exception: subnational empirical analysis suggests higher overall development associated with smaller negative effect of temperature increases, implying advanced countries’ “hot” regions are better insulated.
- Policy recommendations and priorities:
  - Strengthen fiscal buffers (low public debt, build fiscal space).
  - Improve public investment efficiency and public sector governance to ensure transfers translate into productive infrastructure.
  - Facilitate factor reallocation (financial liberalization, reduce bureaucratic impediments) to speed recovery.
  - Preserve human capital (public works and human-capital-preserving programs) to limit hysteresis.
  - Public support for private adaptation investment (subsidies) can be cost-effective (each $1 can yield $2 aggregate damage reduction over 20 years in model).
  - Greater international commitment to climate change mitigation is needed to limit future temperature increases; supporting low-income countries in adaptation is both moral duty and sound global economic policy.

*Source: wpiea2019178-print-pdf (IMF staff; Notre Dame Global Adaptation Index).*

### 2015. University of Notre Dame Global Adaptation Index. Country Index

### 2015. University of Notre Dame Global Adaptation Index. Country Index

### Key literature and empirical foundations
- Costinot, Arnaud, Dave Donaldson, and Cory Smith. "Evolving comparative advantage and the impact of climate change in agricultural markets: Evidence from 1.7 million fields around the world." Journal of Political Economy 124.1 (2016): 205-248.
- Dell, Melissa, Benjamin F. Jones, and Benjamin A. Olken. 2012. “Temperature Shocks and Economic Growth: Evidence from the Last Half Century.” American Economic Journal: Macroeconomics 4 (3): 66–95.
- Deryugina, Tatyana. 2011. “The Role of Transfer Payments in Mitigating Shocks: Evidence from the Impact of Hurricanes.” MPRA Paper 53307.
- Deryugina, Tatyana, and Solomon M. Hsiang. 2014. “Does the Environment Still Matter? Daily Temperature and Income in the United States.” NBER Working Paper 20750.
- Deschênes, Olivier, and Michael Greenstone. 2011. “Climate change, mortality, and adaptation: Evidence from annual fluctuations in weather in the US.” American Economic Journal: Applied Economics 3.4: 152-85.
- Farid, Mai, Michael Keen, Michael Papaioannou, Ian Parry, Catherine Pattillo, and Anna Ter-Martirosyan. 2016. “After Paris: Fiscal, Macroeconomic, and Financial Implications of Climate Change.” IMF Staff Discussion Note 16/01.
- Hallegatte, Stéphane. 2009; 2011; 2016. Key works on adaptation strategies, policy design, and poverty impacts.
- Hsiang, Solomon M., and Daiju Narita. 2012; Hsiang and Amir Jina. 2014. Adaptation to cyclone risk and long-run growth effects from cyclones (NBER Working Paper 20352).
- Additional cited works include Heal and Park (2013), Huq et al. (2004), Kahn (2005), Klein Goldewijk et al. (2016), Lane and Milesi-Ferretti (2017), Ilzetzki, Reinhart, and Rogoff (2008), IMF (2016a, 2016b, 2017), IPCC (2014), and others listed in the source.

### Annex 1 — Data sources and country groupings
- Annex Table 1. Data Sources (indicator -> source):
  - Temperature and Precipitation, Historical (Grid Level) -> University of East Anglia, Climate Research Unit (CRU TS v.3.24)
  - Population 1950 (Grid Level) -> History Database of the Global Environment (HYDE v3.2); Klein et al. (2016)
  - Real GDP per Capita -> IMF, World Economic Outlook database; World Bank, World Development Indicators database
  - Subnational GDP per Capita -> Gennaioli et al. (2014)
  - Consumer Price Index -> IMF, World Economic Outlook database
  - Debt-to-GDP Ratio -> IMF, Historical Public Debt Database
  - Reserves Minus Gold -> Lane and Milesi-Ferretti (2017), External Wealth of Nations database updated to 2015
  - Net Official Development Assistance and Official Aid Received -> World Bank, World Development Indicators database
  - Personal Remittances Received -> World Bank, World Development Indicators database
  - Exchange Rate Regime Indicator -> Reinhart and Rogoff (2004); Ilzetzki, Reinhart, and Rogoff (2008), updated to 2015
  - Adaptation Readiness and Capacity -> Notre Dame Global Adaptation Initiative (ND-GAIN), Chen et al. (2015)
  - Domestic Financial Sector Liberalization Index -> Abiad, Detragiache, and Tressel (2008)
  - Quinn-Toyoda Capital Control Index -> Quinn (1997); Quinn and Toyoda (2008)
  - Human Capital Index -> Penn World Tables 9.0
  - Paved Roads Kilometers per Capita -> Calderón, Moral-Benito, and Servén (2014); World Bank, World Development Indicators database; International Road Federation, World Road Statistics
  - Revised Combined Polity Score (Polity2) -> Polity IV / Transparency International
  - Gini Coefficient -> Standardized World Income Inequality Database

- Annex Table 2. Country and Territory Groups (as listed in source):
  - Advanced Economies: Australia, Austria, Belgium, Canada, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hong Kong SAR, Iceland, Ireland, Israel, Italy, Japan, Korea, Latvia, Lithuania, Luxembourg, Macao SAR, Malta, Netherlands, New Zealand, Norway, Portugal, Puerto Rico, San Marino, Singapore, Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Taiwan Province of China, United Kingdom, United States
  - Emerging Market Economies: Albania, Algeria, Angola, Antigua and Barbuda, Argentina, Armenia, Azerbaijan, The Bahamas, Bahrain, Barbados, Belarus, Belize, Bosnia and Herzegovina, Botswana, Brazil, Brunei Darussalam, Bulgaria, Cabo Verde, Chile, China, Colombia, Costa Rica, Croatia, Dominica, Dominican Republic, Ecuador, Egypt, El Salvador, Equatorial Guinea, Fiji, Gabon, Georgia, Grenada, Guatemala, Guyana, Hungary, India, Indonesia, Iran, Iraq, Jamaica, Jordan, Kazakhstan, Kosovo, Kuwait, Lebanon, Libya, Macedonia FYR, Malaysia, Maldives, Marshall Islands, Mauritius, Mexico, Micronesia, Montenegro, Morocco, Namibia, Nauru, Oman, Pakistan, Palau, Panama, Paraguay, Peru, Philippines, Poland, Qatar, Romania, Russia, Samoa, Saudi Arabia, Serbia, Seychelles, South Africa, Sri Lanka, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines, Suriname, Swaziland, Syria, Thailand, Timor-Leste, Tonga, Trinidad and Tobago, Tunisia, Turkey, Turkmenistan, Tuvalu, Ukraine, United Arab Emirates, Uruguay, Vanuatu, Venezuela
  - Low-Income Developing Countries: Afghanistan, Bangladesh, Benin, Bhutan, Bolivia, Burkina Faso, Burundi, Cambodia, Cameroon, Central African Republic, Chad, Comoros, Democratic Republic of the Congo, Republic of Congo, Côte d'Ivoire, Djibouti, Eritrea, Ethiopia, The Gambia, Ghana, Guinea, Guinea-Bissau, Haiti, Honduras, Kenya, Kiribati, Kyrgyz Republic, Lao P.D.R., Lesotho, Liberia, Madagascar, Malawi, Mali, Mauritania, Moldova, Mongolia, Mozambique, Myanmar, Nepal, Nicaragua, Niger, Nigeria, Papua New Guinea, Rwanda, Senegal, Sierra Leone, Solomon Islands, Somalia, South Sudan, Sudan, São Tomé and Príncipe, Tajikistan, Tanzania, Togo, Uganda, Uzbekistan, Vietnam, Yemen, Zambia, Zimbabwe
  - Countries and Territories with Average Annual Temperature above 15°C: (full list as in source)
  - Countries with Province-Level Data: (full list as in source)
- Note: Source states "* Not included in the main regression analysis." for countries marked with an asterisk.

### Annex 2 — Modeling optimal adaptation (theoretical framework and results)
- Extension of DIG model to include private adaptation and public subsidies to private adaptation; damages modeled as before.
- In absence of adaptation, increased temperature causes gross damage at time t in sector j:
  - ( ) f jt jt jt GD gdT q ==
- Firm i’s capacity to adapt to climate change denoted by ,i jt O . It is increasing in firm i’s protection expenditures ,i jt AD and in total sectoral protection expenditures 1 0, d jt i jt ADADi= ∫ .
- Residual damage for firm i in sector j:
  - , , ( ,) jt i jt jt i jti jt gd OADAD φ Ω =
  - The marginal damage reduction from adaptation spending is decreasing. The positive parameter φ is the elasticity of damage reduction to the level of adaptation.
- If cost of a unit of protection is equal to , AD t P and the functional form for capacity to adapt is
  - () ,, ,; jtjt i jti jti jt OADADAD AD ς ς =
  - (with 01ς≤≤ ), then cost minimization by firms in symmetric equilibrium , jt i jt ADAD = determines the optimal level of adaptation expenditure for each firm:
  - ( ) 1 11 , , . jt i jt AD t GD AD P φς φ ++  =   
- The optimal level of firm-specific residual damage is then:
  - ( ) 1 , jt jt jt gd AD φς+ Ω =
  - This firm-level optimum can be shown to be socially suboptimal.
- Social planner’s cost function differs from individual firms:
  - () ( ) 1 , , SPSPSP i jt jtjtAD tjt TotDGDADPAD φς−+ = +
- Minimizing social cost gives socially optimal adaptation expenditures:
  - ( ) ( ) 1 11 , 1. jt SP jt AD t GD AD P φς φς ++  = +   
- The adaptation spending gap (as a fraction of the socially optimal adaptation spending):
  - ( ) 1 11 1 1 1 φς ς ++  −  + 
- Government subsidy required per unit cost of protection to achieve social optimum:
  - It can be shown that the socially optimal amount of adaptation expenditures can be achieved if subsidies in the amount of ,jtς υ per unit cost of protection are paid by the government to the firms
  - () , 1 jtς ς υ ς = +

*Source: Authors’ compilation from "2015. University of Notre Dame Global Adaptation Index. Country Index", Technical Report.*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019178-print-pdf.pdf_