## wpiea2022237-print-pdf

## Source details

**Canonical URL:** [wpiea2022237-print-pdf](https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022237-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2022/english/wpiea2022237-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2022/english/wpiea2022237-print-pdf.pdf.json)

---

### Overview of climate damages and event frequency
- Five episodes recorded damages exceeding 10 percent of GDP, including storms Fifi and Mitch with damages of 33 percent in Honduras in 1974 and around 60 percent in 1998.
- According to EM-DAT, reported intensity and frequency of climate events in the region are larger after 2000.
- Based on available monthly economic activity data (2000-2019), events in the sample period have been less severe on average: 2.54 vs 0.97 percent of GDP.
- Total damages as defined by EM-DAT correspond to damage value at the moment of the event and are available only for about 1/3 of geological disasters.
- 6.4 percent.
- 6.4 percent.

### Limitations of annual-data studies and motivation for high-frequency analysis
- Extreme climate events tend to be short-lived; annual data make it difficult to accurately estimate direct impacts:
  - Average duration of climate events (storms and floods) is 0.12 months.
  - 83.6 percent of events have an average duration of less than 6 days.
  - Around 40 percent of climate events occurred in the last quarter of the year (25.4 percent in October and 14.3 percent in November-December).
  - More than 5 percent of climate events occurred in the first quarter of the year.
- Economic responses (reconstruction, insurance payouts, family transfers, government support) can offset impacts at annual frequency, creating bias from endogenous policy responses.
- Annual data can dilute identification when periods of high precipitation are followed by lower-than-average precipitation in the remainder of the year.
- Prior annual-frequency estimates for CAPDR show storms and floods reduce real GDP growth by around 1 percentage point (IMF (2017)).

### Literature context and complementary high-frequency approaches
- Two strands:
  - Effects of temperature/precipitation changes on economic activity (e.g., Burke et al., 2015; Bansal and Ochoa, 2011).
    - Hsiang (2010): twenty-eight Caribbean-basin countries lose national output by 2.5 percent to 1°C warming.
    - Dell, Jones, and Olken (2012): 1°C increase in temperature drops GDP per capita growth for poor countries.
  - Extreme climate events using EM-DAT event-level economic losses (e.g., Raddatz, 2009; Acevedo, 2014).
- Recent high-frequency approaches:
  - Akyapi et al. (2022): construct over 160 climate variables from high-frequency data; find larger share of GDP variation explained by weather variation than previously estimated.
  - Kim et al. (2021): use Actuaries Climate Index (ACI) to assess monthly impacts of severe weather on US economic activity.

### Methodology and empirical contributions
- First application exploiting monthly high-frequency macroeconomic data in the CAPDR region to estimate impacts of monthly climate events.
- Identification and model:
  - EM-DAT used to identify dates of tropical storms and floods; create dummy variables (1 = event, 0 = otherwise); months with multiple events can have dummy values greater than 1.
  - Estimate a Vector Autoregression (VAR) at monthly frequency (typically 2000-2019), including climate dummies, domestic activity variables, controlling for US activity and oil prices.
  - Benchmark model: country-specific VARX with y_t vector of five endogenous variables (including climate variable) and z_t vector of two exogenous variables (US GDP growth and oil prices).
  - Identification via Cholesky decomposition with climate dummy ordered first and contemporaneously exogenous.
- Data coverage:
  - Countries: Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, Panama, Dominican Republic.
  - Monthly variables: index of economic activity (IMAE), bilateral exchange rate with US dollar (where applicable), consumer price index, remittances.
  - Use seasonally unadjusted data and year-over-year growth rates for all variables.
  - Data covers 2001 to 2019; 2020 excluded due to Covid pandemic effects (storms Eta and Iota).
  - Costa Rica exception: IMAE series goes back to 1991.
- Caveats:
  - Fiscal data and interest rate data not long enough for many countries and thus not included.
  - Creating dummies assigns equal weight to all disaster types, potentially ignoring heterogeneous magnitudes.

### Key empirical findings
- IMAE growth impacts (country-specific):
  - Monthly year-over-year economic activity decreases by around 0.5-1 percentage points in affected Central American countries.
  - Dominican Republic: economic growth contracts by around 0.5 of a percentage point on impact.
  - El Salvador: economic growth contracts by 1 percentage point on impact.
  - Honduras: monthly economic activity declines by around 0.75 percentage points around three to five months after an event.
  - Costa Rica: monthly economic activity declines by around 0.75 percentage points around three to five months after an event.
  - Nicaragua: output declines but impact is not statistically significant.
  - Panama: output increases on impact; impact statistically insignificant.
  - Interpretation: output growth returns to long-run average but not to pre-climate shock trend; persistent level effects consistent with Hsiang and Jina (2014) and Akyapi et al. (2022).
- Remittances:
  - Honduras: remittances increase by around 3 percentage points in response to extreme climate events.
  - Guatemala: remittances increase by around 8 percentage points in response to extreme climate events.
  - Dominican Republic, El Salvador, Nicaragua: responses are not statistically significant.
  - In VAR estimates, an exogenous remittances shock has a positive and statistically significant impact on economic activity in all cases.
  - Remittances as mitigation: remittances comprise a larger share of GDP in Honduras than Guatemala: 24 percent of GDP vs 16 percent.
- Additional macro impacts:
  - Inflation: for most countries climate events lead to a statistically significant increase in inflation by 0.2- ½ p.p., with peak impact 2 to 5 months after the shock.
  - Exchange rate: appreciation in countries that received sizable remittances inflows (Honduras and Guatemala); otherwise effects are insignificant.
- Panel VAR and aggregated impacts:
  - Panel VAR on impact: year-over-year IMAE growth declines between 0.2 and 0.6 pp.
  - Year-over-year remittances increase by more than 1.5 pp on impact.
  - Storms have greater impact than floods on both economic activity and remittances; storms’ impact on IMAE tends to be longer lasting with a peak after two months.
  - Annualizing estimated IMAE impacts from extreme storm events for the region implies a decrease of 0.6 percentage points in output.
  - Prior comparable findings: IMF (2017) finds storms and floods contract GDP growth by around 1 p.p. over 1950-2016; Bello (2017) finds impact on per capita GDP growth of around ¾ p.p. on impact from storms and around ¼ p.p. for other climate events.

### Index construction and augmented-specification robustness (Honduras and Guatemala)
- Climate index construction:
  - Index follows spirit of ACI; country indices aggregate five components after dropping sea level due to missing data.
  - Components (as defined):
    1. T90 (extreme high temperature): frequency change of temperatures above the 90th percentile relative to the reference period of 1961 to 1990.
    2. T10 (extreme low temperature): frequency change of temperatures below the 10th percentile relative to the reference period of 1961 to 1990.
    3. Rx5days (heavy precipitation): maximum five-day rainfall in the month
    4. CDD (drought): maximum number of consecutive days with less than 1mm of daily precipitation
    5. W (high wind): frequency change of daily wind power above the 90th percentile relative to the reference period of 1961 to 1990.
  - Each component standardized relative to the reference period (1961-1990); ACI defined as simple average of standardized components.
  - For Honduras and Guatemala: temperatures and precipitations from Climdex; daily wind speed from IOWA Environmental Mesonet (IEM) converted; sea level dropped; T10 sign changed to positive so cold extremes increase the aggregated index.
- Augmented specification:
  - Multiply climate index with EM-DAT climate dummy to approximate event intensity.
  - One standard deviation shock in intensity yields significant increase in remittance growth and significant decrease in IMAE growth.
  - Sectoral disaggregation (IMAE subsectors):
    - Primary (agriculture and mining) drops about 1 percentage points driving decrease in aggregate IMAE growth.
    - Secondary (manufacturing and construction) response unclear first few months but barely drops and becomes significant after 5 months.
    - Tertiary (services and residuals) increases on impact, possibly due to increased health expenditures or government supports.
  - Augmented specification produces somewhat larger and more significant effects, especially for Honduras.
  - Converted annual impacts: -0.6 percentage points for Honduras; 0.2 percentage points for Guatemala. This implies that the annual effect on economic growth is roughly equal to or twice as severe as the monthly effect.

### Policy implications and recommendations
- More accurate high-frequency estimates are key for policy makers to better design policies, including:
  - Ensuring the availability of sufficient resources to respond to each climate event.
  - Designing adaptation policies to mitigate future impacts.
- Incorporating climate risks into macro-fiscal frameworks is recommended, for instance through climate scenarios that incorporate temporary weather disasters to determine the “adequate size of buffers and risk-reduction strategies”.
- Given limited fiscal resources and competing needs, CAPDR countries should carefully allocate resources and consider distributional effects of climate adaptation policies; cost-benefit analysis (CBA) should be an integral tool for evaluation.
- Future work: explore non-linearities of extreme climate events exploiting the newly constructed ACI index.

### Key statistics and exact figures (preserved verbatim)
- 0.2 percentage points for Guatemala. This implies that the annual effect on economic growth is
- Monthly year-over-year economic activity decreasing by around 0.5-1 percentage points.
- Remittances increasing by more than 3 percentage points on impact in the case of Guatemala and Honduras.
- Study period: 2000 to 2019.
- Reference period for percentiles: 1961 to 1990.
- ACI originally aggregates six components; the constructed country index aggregates five indicators after dropping sea level.

*wpiea2022237-print-pdf - Introduction.*

### 6.4 percent.

### 6.4 percent.

### Overview of climate damages and event frequency
- Five episodes recorded damages exceeding 10 percent of GDP, including storms Fifi and Mitch with damages of 33 percent in Honduras in 1974 and around 60 percent in 1998.
- According to EM-DAT, reported intensity and frequency of climate events in the region are larger after 2000.
- Based on available monthly economic activity data (2000-2019), events in the sample period have been less severe on average: 2.54 vs 0.97 percent of GDP.
- Total damages as defined by EM-DAT correspond to damage value at the moment of the event and are available only for about 1/3 of geological disasters.

### Limitations of annual-data studies and motivation for high-frequency analysis
- Extreme climate events tend to be short-lived; annual data make it difficult to accurately estimate direct impacts.
  - Average duration of climate events (storms and floods) is 0.12 months.
  - 83.6 percent of events have an average duration of less than 6 days.
  - Around 40 percent of climate events occurred in the last quarter of the year (25.4 percent in October and 14.3 percent in November-December).
  - More than 5 percent of climate events occurred in the first quarter of the year.
- Economic responses (reconstruction, insurance payouts, family transfers, government support) can offset impacts at annual frequency, creating bias from endogenous policy responses.
- Annual data can dilute identification when periods of high precipitation are followed by lower-than-average precipitation in the remainder of the year.
- Prior annual-frequency estimates for CAPDR show storms and floods reduce real GDP growth by around 1 percentage point (IMF (2017)).

### Literature context and complementary high-frequency approaches
- Two strands:
  - Effects of temperature/precipitation changes on economic activity (e.g., Burke et al., 2015; Bansal and Ochoa, 2011).
    - Hsiang (2010): twenty-eight Caribbean-basin countries lose national output by 2.5 percent to 1°C warming.
    - Dell, Jones, and Olken (2012): 1°C increase in temperature drops GDP per capita growth for poor countries.
  - Extreme climate events using EM-DAT event-level economic losses (e.g., Raddatz, 2009; Acevedo, 2014).
- Recent high-frequency approaches:
  - Akyapi et al. (2022): construct over 160 climate variables from high-frequency data; find larger share of GDP variation explained by weather variation than previously estimated.
  - Kim et al. (2021): use Actuaries Climate Index (ACI) to assess monthly impacts of severe weather on US economic activity.

### Methodology and empirical contributions
- First application exploiting monthly high-frequency macroeconomic data in the CAPDR region to estimate impacts of monthly climate events.
- Identification:
  - Use EM-DAT to identify dates of tropical storms and floods; create dummy variables (1 = event, 0 = otherwise).
  - Estimate a Vector Autoregression (VAR) at monthly frequency (typically 2000-2019), including climate dummies, domestic activity variables, controlling for US activity and oil prices.
- Data coverage:
  - Countries: Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, Panama, Dominican Republic.
  - Monthly variables: index of economic activity (IMAE), bilateral exchange rate with US dollar (where applicable), consumer price index, remittances.
  - Use seasonally unadjusted data and year-over-year growth rates for all variables.
  - Data covers 2001 to 2019; 2020 excluded due to Covid pandemic effects (storms Eta and Iota).
  - Costa Rica exception: IMAE series goes back to 1991.
- Caveats:
  - Fiscal data and interest rate data not long enough for many countries and thus not included.
  - Creating dummies assigns equal weight to all disaster types, potentially ignoring heterogeneous magnitudes.

### Key empirical findings
- Climate events in Central American countries decrease year-over-year GDP growth by about 0.5-1 percentage points.
  - These estimates are in line with or higher than previous annual-data estimates for the region.
- Remittances or transfers from nationals living abroad increase by more than 3 percentage points in response to an extreme climate event in Central American countries, consistent with remittances acting as a shock absorber or private insurance mechanism.

### Index construction and robustness checks
- Construct an extreme weather index following the spirit of the Actuaries Climate (ACI) Index to better quantify intensity of events at higher frequency.
  - Index constructed for Honduras and Guatemala, where long time series of daily (monthly) weather data for temperature, wind, and precipitation exist.
  - For other CAPDR countries, weather data are available but often interpolated from neighboring stations, which reduces quality.
- Using this index to control for severity of events shows results are robust, implying climate event dummies capture average intensity reasonably well.
- Suggestion for future work: explore non-linearities of extreme climate events exploiting the newly constructed ACI index.

### Organization of the paper and data appendix
- Remaining structure: Section 2 describes data and empirical model; Section 3 provides results; Section 4 concludes; Appendix contains further information about the climate index and additional results.

*wpiea2022237-print-pdf - 6.4 percent.*

### Introduction.

### Introduction

### Data limitations and scope
- Quarterly data is not available over a long time period in the region; the relatively short sample period does not lend itself to examine the impact of climate change.
- The short sample period is noted as an advantage because it allows detracting from looking at climate trends (see Akyapi et al. (2022) mentioned in the source).

### Role of remittances and exchange rates in estimation
- All VAR estimates, other than Costa Rica and Panama (who send more remittances than they receive and do not have data at the monthly frequency), include remittances to examine their responses and whether remittances play a key supporting role.
- All VAR estimates excluding El Salvador and Panama, countries with fixed exchange rates, include the bilateral exchange rate to capture the importance of the open economy channel.
- Two exogenous variables are included to control for external factors: monthly US GDP and brent oil prices.
  - Monthly GDP is constructed by S&P based on the construction of NIPA quarterly national accounts data.
  - Monthly measures of industrial production were also considered; results do not change significantly.

### Climate events data and constructed index
- Two alternative measures of extreme climate events are used:
  - Monthly dummy variables of flood and storm within each CAPDR country using the EM-DAT database, assigning a 1 to a reported tropical storm or flood. More specifically, each event is recorded as 1 and months with multiple events can have dummy values greater than 1 (e.g., storm and flood in same month → dummy records value 2).
  - A monthly climate index based on the ACI that provides information about severe weather conditions (Appendix 1).
- The climate index is constructed for Honduras and Guatemala (the two available countries in terms of data availability and quality).
  - Each country’s index is an aggregation of 5 weather components: high temperatures, low temperatures, heavy rainfall, drought, and high wind.
  - Each component is standardized relative to the reference period (1961-1990).
  - The reference period is set by construction of some indices collected from Climdex; the ACI is based on the same reference period.
  - Each of the series underlying the index are presented in Appendix A.
  - The ending period of each of the two series is due to unreliable data quality afterwards.
- Interpretation of index values:
  - Positive (red) values imply that weather has become more extreme since 1990.
  - In Figure 2 (Honduras 1973-2017; Guatemala 1973-2018), red bars start to dominate from mid-1990 indicating that extreme weather has increased over time.
  - The climate index largely corresponds to reported climate events in EM-DAT such as storms (green star) and floods (black circle).
  - Despite greater intensity of climate events since 2000, reported damages are lower compared to 1960-2000 events.
  - Bello (2017) reports similar results for Latin America and the Caribbean.

### Empirical model (benchmark VARX)
- Benchmark econometric model: country-specific vector autoregression with exogenous variables (VARX):
  - y_t = c + Φ_1 y_{t-1} + ... + Φ_p y_{t-p} + Γ_0 z_t + Γ_1 z_{t-1} + ... + Γ_q z_{t-q} + B ε_t
  - y_t is a vector of five endogenous variables including the climate variable (either dummy variable or climate index times the dummy) and domestic economic variables: year-over-year IMAE growth, inflation, exchange rate and remittances.
  - z_t is a vector of two exogenous variables: US GDP growth and oil prices.
  - p and q represent lag lengths for the endogenous and exogenous variables, respectively; lag length criteria tests used for endogenous variables; for exogenous variables include either the first or second lag depending on explanatory power.
- Identification:
  - Zero contemporaneous restrictions (Cholesky decomposition) assume B is a lower triangular matrix.
  - Ordering in y_t: first the climate dummy and the climate index times the dummy.
  - Restrictions imposed so that all economic variables do not have an impact on the climate variable in contemporaneous terms (first row of each matrix: all coefficients apart from that of the lagged climate event set equal to zero).
  - This implies climate shock is entirely exogenous in the short term; economic variables can interact with lagged climate disaster variable and country characteristics.
- Alternative estimation methods and robustness:
  - Panel estimates tried but country-specific VAR set as benchmark to see differences between countries; panel results shown as robustness analysis.
  - Local projections (Jordà (2005)) also considered; results robust and available on request.
  - Results for lag-selection and other details are available on request.

### Results — overview and inference
- Impulse response functions (IRFs) to a one-time climate shock are presented with 68 percent bootstrap confidence intervals.
- Baseline specification: country-specific VARs using the “dummy based” climate variable (storm and flood combined).
  - The dummy variable response is normalized to 1 on impact.
  - Focus on IMAE growth and remittance growth; full country results in Appendix B.1.

### Baseline results — IMAE growth (country-specific)
- Responses of IMAE (year-over-year) growth to a climate event:
  - Dominican Republic: economic growth contracts by around 0.5 of a percentage point on impact.
  - El Salvador: economic growth contracts by 1 percentage point on impact.
  - Honduras: monthly economic activity declines by around 0.75 percentage points around three to five months after an event.
  - Costa Rica: monthly economic activity declines by around 0.75 percentage points around three to five months after an event.
  - Nicaragua: output declines but impact is not statistically significant.
  - Panama: output increases on impact; impact statistically insignificant.
- Interpretation:
  - Results suggest while output growth returns to its long-run average, there is no return to the pre-climate shock trend.
  - Insignificant impact in Nicaragua and Panama partly due to relatively small climate events during estimation period as measured by damages in EM-DAT (Table 1b).
  - Persistent level effects are consistent with findings in Hsiang and Jina (2014) and Akyapi et al. (2022).

### Baseline results — remittances (country-specific)
- Remittance growth responses to extreme climate events:
  - Honduras: remittances increase by around 3 percentage points in response to extreme climate events.
  - Guatemala: remittances increase by around 8 percentage points in response to extreme climate events.
  - Dominican Republic, El Salvador, Nicaragua: responses are not statistically significant.
- Context and interpretation:
  - Costa Rica and Panama send more remittances than they receive.
  - The increase in remittances in Honduras and Guatemala is consistent with a mitigation role of remittances acting as private insurance for consumption smoothing.
  - The response is characterized as a compassionate effect: immigrant workers in the US send more transfers to cushion impacts.
  - Hondurans and Guatemalans are more recent migrants into the US relative to other countries considered, implying stronger ties and larger first-generation immigrant shares.
  - Remittances comprise a larger share of GDP in Honduras than Guatemala: 24 percent of GDP vs 16 percent.
  - For the Dominican Republic, Nicaragua, and El Salvador many migrants have been in the US for prolonged periods or are US nationals; statistics cited (Pew) note that by 2017 around 43% of foreign-born Dominicans had been in the US for at least 20 years and 53% were US nationals; for Nicaraguans 64% had been living in the US for over 20 years and 61% were US citizens; for Salvadorians 44% had been living in the US for over 20 years and 33% were US citizens.
  - In VAR estimates, an exogenous remittances shock has a positive and statistically significant impact on economic activity in all cases.

### Additional macro impacts
- Inflation:
  - For most countries climate events lead to a statistically significant increase in inflation by 0.2- ½ p.p., with peak impact 2 to 5 months after the shock, consistent with destruction of agricultural production.
- Exchange rate:
  - Results point to an appreciation in countries that received sizable remittances inflows (Honduras and Guatemala); otherwise effects are insignificant.

### Robustness checks — panel VAR estimates
- Panel VAR results (stacking country observations) broadly similar to country-specific results:
  - On impact, year-over-year IMAE growth declines between 0.2 and 0.6 pp.
  - Year-over-year remittances increase by more than 1.5 pp on impact.
  - Storms have greater impact than floods on both economic activity and remittances; storms’ impact on IMAE tends to be longer lasting with a peak after two months.
- Annualizing estimated IMAE impacts from extreme storm events for the region:
  - High-frequency monthly panel estimation implies a decrease of 0.6 percentage points in output.
  - Comparable literature findings: IMF (2017) finds storms and floods contract GDP growth by around 1 p.p. over the sample 1950-2016; Bello (2017) finds impact on per capita GDP growth of around ¾ p.p. on impact from storms and around ¼ p.p. for other climate events.
  - Similarity of results with earlier studies suggests that impacts of extreme climate effects might have been underestimated using annual data in earlier periods.

### Robustness checks — augmented results using the monthly climate index (Honduras and Guatemala)
- Motivation and advantages of augmenting dummies with the climate index:
  - Index measured with physical and meteorological observations to alleviate measurement problems.
  - Higher frequency (daily → monthly) aligned with economic data, avoiding underestimation of true impact.
  - Allows measurement of severity of weather conditions related to identified events.
- Method:
  - Multiply the climate index with the EM-DAT climate dummy to approximate intensity of each disaster.
- Results:
  - One standard deviation shock in intensity of extreme climatic event (Honduras and Guatemala) yields:
    - Significant increase in remittance growth.
    - Significant decrease in IMAE growth.
  - Disaggregation of IMAE into subsectors (primary, secondary, tertiary) shows:
    - Primary (agriculture and mining) drops about 1 percentage points driving decrease in aggregate IMAE growth.
    - Secondary (manufacturing and construction) response unclear first few months but barely drops and becomes significant after 5 months.
    - Tertiary (services and residuals) increases on impact, possibly due to increased health expenditures or government supports.
- Comparison with dummy-only specification:
  - Dynamics of responses are similar between country-specific VARs using dummies and augmented specification using ACI index multiplied by dummies.
  - Augmented specification produces somewhat larger and more significant effects, especially for Honduras, providing a better estimate of outcomes associated with climate events.
- Converted annual impacts from the climate index specification:
  - Monthly impact converted to annual growth impact yields -0.6 percentage points for Honduras (the text cuts off for Guatemala’s converted annual impact in the provided excerpt).

*Source: wpiea2022237-print-pdf - Introduction.*

### 0.2 percentage points for Guatemala. This implies that the annual effect on economic growth is

### wpiea2022237-print-pdf - 0.2 percentage points for Guatemala. This implies that the annual effect on economic growth is

### IV. Final Remarks — Core findings
- The paper estimates the impact of extreme climate events using high-frequency data for seven Central American countries: Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, Panama, and the Dominican Republic.
- The analysis exploits monthly dummy variables and country-specific VAR and panel VAR with exogenous variables, covering the period from 2000 to 2019.
- Findings indicate that climate events result in statistically significant macroeconomic impact in most Central American countries:
  - Monthly year-over-year economic activity decreases by around 0.5-1 percentage points.
  - Remittances increase by more than 3 percentage points on impact in the case of Guatemala and Honduras.
- The authors note that annual-frequency estimates are likely to underestimate the true effects of extreme climate events because:
  - Effects may be short-lived and thus obscured in annual aggregates.
  - Reconstruction or external/government support may offset direct impacts within the year.

### Climate index construction (Appendix A) — methodology and components
- The climate index for Honduras and Guatemala is constructed following the spirit of the Actuaries Climate Index (ACI).
- ACI aggregates six weather components (as defined in the source):
  1. T90 (extreme high temperature): frequency change of temperatures above the 90th percentile relative to the reference period of 1961 to 1990.
  2. T10 (extreme low temperature): frequency change of temperatures below the 10th percentile relative to the reference period of 1961 to 1990.
  3. Rx5days (heavy precipitation): maximum five-day rainfall in the month
  4. CDD (drought): maximum number of consecutive days with less than 1mm of daily precipitation
  5. W (high wind): frequency change of daily wind power above the 90th percentile relative to the reference period of 1961 to 1990.
  6. S (sea level): change in the sea level
- To aggregate components, each variable is standardized relative to the reference period. The ACI is defined as a simple average of the standardized components. An increased value of the index indicates that the incidence of severe weather has increased.
- For Honduras and Guatemala, data sources and modifications:
  - Temperatures and precipitations from Climdex.
  - Daily wind speed from IOWA Environmental Mesonet (IEM) converted to the standardized high wind variable.
  - Sea level component dropped due to many missing observations; the constructed index aggregates five weather indicators (i.e., the six-component ACI minus S).
  - For Honduras and Guatemala, T10 showed an increasing and cyclical trend after mid-1980s; the authors changed the T10 sign to positive in the aggregation so that an increase of cold extremes increases the aggregated index.
- Monthly index series and component plots are presented in figures (Figures A1, A2, and main text Figures 2 and 3).

### Robustness and specification notes
- The paper compares IRFs between specifications using dummy variables and an index-weighted approach (index multiplied by a dummy so that each event can be weighted differently).
- Results using the constructed monthly extreme climate index for Honduras and Guatemala are qualitatively similar to estimates using dummy variables.
- The authors note that dummy-variable approaches do not exploit heterogeneity of different events because dummies assign the same weight to different climate events.

### Policy implications and recommendations
- More accurate estimates of the impact of climate effects are key for policy makers to better design policies, including:
  - Ensuring the availability of sufficient resources to respond to each climate event.
  - Designing adaptation policies to mitigate future impacts.
- Incorporating climate risks into macro-fiscal frameworks is recommended, for instance through relevant climate scenarios that incorporate temporary weather disasters to determine the “adequate size of buffers and risk-reduction strategies”.
- Given limited fiscal resources and competing needs, CAPDR countries should carefully allocate resources and consider distributional effects of climate adaptation policies; cost-benefit analysis (CBA) should be an integral tool for evaluation.
- Future work should explore non-linearities of extreme climate events exploiting the newly constructed ACI index.

### Key statistics and exact figures (preserved verbatim)
- 0.2 percentage points for Guatemala. This implies that the annual effect on economic growth is roughly equal to or twice as severe as the monthly effect.
- Monthly year-over-year economic activity decreasing by around 0.5-1 percentage points.
- Remittances increasing by more than 3 percentage points on impact in the case of Guatemala and Honduras.
- Study period: 2000 to 2019.
- Reference period for percentiles: 1961 to 1990.
- ACI originally aggregates six components; the constructed country index aggregates five indicators after dropping sea level.

*IMF Working Paper (content unit: wpiea2022237-print-pdf).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022237-print-pdf.pdf_
