## How Persistent are Climate-Related Price Shocks — Excerpts (wpiea2022207-print-pdf)

## Source details

**Canonical URL:** [How Persistent are Climate-Related Price Shocks — Excerpts (wpiea2022207-print-pdf)](https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022207-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2022/english/wpiea2022207-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2022/english/wpiea2022207-print-pdf.pdf.json)

---

### Introduction — context, objective, and high-level findings
- Context and motivation:
  - Climate change increases frequency and severity of supply and demand shocks, complicating monetary policy formulation.
  - Extreme weather events (e.g., droughts) increase food price volatility; adverse supply shocks can push prices up and output down.
  - Policy uncertainty about climate-related transition policies could depress potential output and lower the equilibrium real interest rate.
- Objective and scope:
  - Main objective: investigate how climate shocks affect consumer prices across a broad range of countries.
  - Holistic approach: natural disasters and temperature/precipitation shocks on headline and food inflation.
  - Coverage: up to 183 countries over 1970-2018 at quarterly frequency.
- Key high-level findings:
  - Frequency and intensity of climate shocks increased over time, driven particularly by storms and floods.
  - Clear evidence of global warming; little evidence of increased precipitation.
  - Increased frequency of natural disasters far higher in developing countries than in advanced and emerging market economies.
  - Droughts tend to have the highest overall positive impact on inflation (especially severe droughts); effects more pronounced for food inflation.
  - Floods tend to have a dampening impact on inflation (demand shocks preeminent).
  - Inflation-targeting (IT) countries tend to experience lower and less persistent impacts from climate shocks.
  - Robustness checks (long-run deviations of climate indicators) generally confirm results with nuances.

### Data and stylized facts (data sources, construction, and summary statistics)
- Data sources and construction:
  - Natural disasters: EM-DAT (CRED, Université Catholique de Louvain).
  - Temperature and precipitation: CRU TS v. 4.03 (University of East Anglia).
  - EM-DAT inclusion: events with 10+ deaths, 100+ affected, declaration of state of emergency, or need for international assistance.
  - Disaster intensity dummy: equals 1 when (fatalities + 0.3 × affected people) > a*, with moderate disasters using a* = 0.0001 and severe disasters using a* = 0.01.
  - Climate z-scores: country-specific z-scores relative to the past 30 years: T_{i,t}^{z} = (T_{i,t} - T_{i,t}^{ave}) / σ_{i,t}^T; precipitation z-scores defined analogously.
  - CPI: monthly CPI for 183 countries from IMF IFS over 1970-2018, converted to quarterly; year-on-year inflation; inflation restricted to range ±20 percent.
  - Country groups: 33 AEs, 76 EMs, 53 LIDCs. Monetary regime sample: 28 IT and 73 non-IT. Monetary policy effects analyzed on post-1990 data only.
- Frequency and distributional stylized facts:
  - About 31.9 percent of country-quarter observations associated with at least one natural disaster.
  - By disaster type (portion of at least one event occurring): Floods: 9 percent; Storms: 5.2 percent; Droughts: 1.6 percent.
  - Flood events averaged less than 50 each year between 1970-1995; more than tripled between 1996-2018. In roughly 1999–2018: floods ~150 events/year; storms ~100 occurrences/year; droughts less frequent (ranked 7th between 1996 and 2018 but rising).
  - Natural disasters increased most for LIDCs: from once every two years (1970-1995) to twice every year (1996-2018).
- Summary statistics (1970-2018):
  - Inflation overall mean: 5.7 percent.
    - AEs mean inflation: 4.5 percent.
    - EMs mean inflation: 6.0 percent.
    - LIDCs mean inflation: 6.3 percent.
    - Non-Inflation Targeting mean inflation: 6.4 percent.
    - Inflation Targeting mean inflation: 3.3 percent.
  - Temperature z-score overall mean: 0.5 (higher in EMs and LIDCs than in AEs; EMs and LIDCs on average more than 50 percent higher than AEs).
  - Precipitation z-score overall mean: 0.03 (close to zero; distributions skewed right).

### Impact of natural disasters on inflation — heterogeneity by disaster type, country group, and monetary regime
- Droughts:
  - An average drought raises inflation by 0.2 percentage point upon impact (statistically significant at the 95 percent level).
  - Impact lingers at a similar level for the next four quarters but becomes statistically insignificant two quarters after the shock.
  - A moderate or severe drought raises inflation by 3 percentage points within four quarters.
  - Impact of drought on inflation is temporary, lasting less than a year.
- Floods:
  - Average floods exert a deflationary reaction (statistically insignificant even at 68 percent level).
  - Severe floods trigger a strong and statistically significant response (95 percent level), reaching 2 percentage points a year after the shock.
  - Empirical negative response consistent with demand shocks (risk-aversion) dominating supply shocks.
- Storms:
  - Initially inflationary; effects short lived.
  - Significant for about two and three quarters for moderate and severe events, respectively.
  - Subsequent negative response remains depressed for more than two years.
  - Magnitude, duration, and significance increase with storm intensity.
- Country-group heterogeneity:
  - Advanced Economies (AEs):
    - All events are inflationary on average.
    - Drought: inflation rises by 2 percentage points in the second quarter, stays elevated about three quarters, then declines (estimates not different from zero even at 68 percent level).
    - Severe floods: strong deflationary impact in first six quarters, decreasing inflation to -5 percentage points within a year, then becoming inflationary and significant the next year.
    - Moderate and severe storms in AEs: have a deflationary impact (contrasting with average storms).
  - Emerging Markets (EMs):
    - Droughts tend to be inflationary.
    - Floods and storms tend to be deflationary.
    - An average flood reduces inflation by < 0.5 percentage point within a year; a moderate flood reduces inflation by 1 percentage point; a severe flood reduces inflation by almost 4 percentage points.
    - An average drought is inflationary; severe droughts could exert deflationary pressures after one year.
  - Low-Income Developing Countries (LIDCs):
    - Similar to EMs: droughts inflationary; floods and storms deflationary. Intensity amplifies magnitude and persistence.
- Monetary regime heterogeneity (IT vs non-IT):
  - General: monetary policy contains inflationary pressures after natural disaster shocks in both AEs and EMs, while it tends to accommodate deflationary impacts.
  - AEs:
    - IT countries: initial drought reaction < 0.5 percentage point; point estimates negative from second quarter onward.
    - Non-IT countries: inflation rises and remains elevated at about 2 percentage points for ~two years.
    - Floods and storms: slightly deflationary in IT AEs; inflationary in non-IT AEs.
  - EMs:
    - IT countries: inflation anchored around zero for droughts and floods; storms increase inflation in second year by 0.1 percentage point.
    - Non-IT countries: drought generates a 3 percentage points jump in inflation; floods increase inflation gradually to 1 percentage point in second year; storms reduce inflation by 1 percentage point in second year.
  - Appendix E: evidence using IAPOC is suggestive but weak that stronger monetary policy frameworks provide insulation.

### Food inflation and subnational/item heterogeneity
- Aggregate food inflation responses:
  - Droughts: food inflation jumps sharply to over 5 percentage points upon impact, rising to nearly 10 percentage points in the third quarter after impact; stays elevated for first year, then reverses negative for about five quarters.
  - Floods: mostly deflationary on food inflation in the first year.
  - Storms: food prices rise for four quarters after storms; effect smaller than for droughts.
- Subnational and item-level findings (Appendix A):
  - Staples (subnational, monthly):
    - Droughts: staples inflation gradually increases for 9 months, then subsides; second year slightly deflationary.
    - Floods: deflationary impact 6-12 months after occurrence.
    - Storms: inflationary and persistent.
    - Magnitudes at subnational level larger and more precisely estimated than national quarterly estimates.
  - Heterogeneity by income:
    - Droughts drive staples inflation in EMs to well above 10 percentage points after ~5 months; impact on LIDCs is limited.
    - Floods’ deflationary impact in 6-9 months more visible for LIDCs.
    - Storms exert inflationary pressures on EMs in first year; on LIDCs in second year.
  - Heterogeneity by food item:
    - Droughts inflationary for staples, meat, fish; deflationary for eggs and milk initially.
    - Floods’ deflationary impact concentrated in sugar.
    - Storms have sizeable deflationary impact on eggs and milk.

### Temporal and spatial mechanisms
- Temporal:
  - Floods tend to be short-lived and often occur during rainy season; replanted crops and harvest timing can mitigate supply effects.
  - Droughts tend to last longer and produce more persistent supply effects.
- Spatial:
  - Floods are spatially concentrated (rivers, lowlands); droughts cover larger areas.
  - Increased precipitation outside immediate flood zones can increase output and induce negative inflationary shocks.

### Robustness, additional controls, and alternative climate measures
- Robustness checks (Appendix B):
  - Extended model includes exchange rate growth, average night lights, and food weight in CPI; results broadly similar to baseline though sample size declines.
  - Notable differences for LIDCs: natural disaster shocks induce inflationary responses in almost all cases except average moderate storms; severe storms are inflationary for just over a year.
- Role of food share in CPI (Appendix C):
  - Median food weight in CPI:
    - Advanced economies: around 15%
    - Emerging markets: 25%
    - Low-income countries: 45%
  - Higher food share leads to larger short-term inflationary pressures following natural disasters.
- Floods and food production (Appendix D):
  - Using FAOSTAT annual data: for rice, maize, and meat, floods on average increase food production in ensuing years; moderate floods tend to be beneficial; severe floods tend to have negative initial impacts.
- Alternative climate measure (Appendix F — SPEI):
  - 3-month SPEI (combines temperature and precipitation) results qualitatively similar to precipitation shocks.
  - Negative SPEI shocks are inflationary for EMs and LIDCs.

### Policy implications and conclusions
- Main implications:
  - Not all climate shocks have similar impacts: droughts—though less frequent—tend to increase inflation; floods more likely to dampen inflation; storms generally inflationary and can be persistent.
  - Droughts’ inflationary impact largely transmitted via food prices.
  - Historical overall impacts muted, but projected increases in frequency and amplitude of climate shocks imply larger impacts over time.
- Monetary policy framework implications:
  - Common frameworks and models are ill-suited for persistent supply shocks; “looking through” supply shocks is inappropriate if they become persistent.
  - Tightening policy in response to climate-driven supply shocks could cause output loss and be less potent if r* declines with climate change.
  - Debate unsettled on alternative frameworks:
    - McKibbin and others (2020): propose nominal income targeting as less demanding.
    - Cantelmo and others (2022): simulations suggest FIT remains superior.
  - Evidence supports that inflation-targeting and stronger monetary frameworks generally better contain inflationary pressures from natural disasters, but some evidence (IAPOC) is weak and heterogeneous.
- Research and operational needs:
  - More research required on appropriate frameworks accounting for heterogeneity by shock type, duration, intensity, and country characteristics.
  - Central banks should consider implications for collateral frameworks, asset purchase programs, and operational feasibility of “green” tilts while balancing monetary policy effectiveness, climate mitigation, and risk protection.

*Source: wpiea2022207-print-pdf — IMF Working Paper excerpts (Introduction; Sections 3, 4.2, 5; Appendices A–F).*

### 1. Introduction ........................................................................................................

### 1. Introduction

### Context and motivation
- Climate change is likely to lead to more frequent and more severe supply and demand shocks that present a challenge to monetary policy formulation.
- Extreme weather events, such as droughts, are expected to increase food price volatility; adverse supply shocks can push prices up and output down (Klomp 2020).
- Frequent supply shocks complicate disentangling permanent from transitory shocks, creating communication and credibility challenges for central banks (NGFS 2020).
- Policy uncertainty about climate-related transition policies could affect demand, investment, and inflation expectations and may depress potential output and lower the equilibrium real interest rate (Bylund and Jonsson 2020).

### Objective and scope of the paper
- Main objective: investigate how climate shocks affect consumer prices across a broad range of countries.
- Holistic approach: examine impacts of different types of natural disasters and temperature and precipitation shocks on headline and food inflation.
- Coverage: a wide variety of countries of differing income status and monetary regimes at national and subnational levels.
- Sample: up to 183 countries over the period 1970-2018 on a quarterly basis.

### Data and empirical approach
- Two sets of climate shock data:
  - Natural disasters: EM-DAT database from the Centre of Research on the Epidemiology of Disasters (CRED) at the Université Catholique de Louvain.
  - Temperature and precipitation: Climate Research Unit of the University of East Anglia (CRU TS v. 4.03).
- Empirical methodology: local projection methods of Jordà (2005) and Jordà and others (2020).
- Focus: not only point estimates but the persistence (duration) of impacts on CPI and components, drawing on Fratzscher and others (2020) and Parker (2018).

### Key findings
- Frequency and intensity:
  - Confirmed that the frequency and intensity of climate shocks have increased over time, driven particularly by storms and floods.
  - Clear evidence of global warming; little evidence of increased precipitation.
  - Increased frequency of natural disasters is far higher in developing countries than in advanced and emerging market economies.
- Heterogeneity by shock type, intensity, country group, and monetary regime:
  - Droughts tend to have the highest overall positive impact on inflation, especially severe droughts, mainly driven by supply shocks.
  - Effects are more pronounced for food inflation.
  - Food typically represents around 40 percent of the CPI basket in LIDCs, amplifying headline inflation effects.
  - Floods tend to have a dampening impact on inflation, suggesting the preeminence of demand shocks (Cantelmo 2022).
  - Monetary policy regime matters: in inflation-targeting (IT) countries climate shocks tend to have lower and less persistent impacts on inflation (consistent with Fratzscher and others 2020).
- Robustness:
  - Results are generally confirmed, with some nuances, when using long-run deviations of climate indicators (less extreme cases of climate shocks).

### Implications for monetary policy
- For LIDCs:
  - Monetary policymaking is particularly challenging due to more significant and prolonged effects on inflation and output.
  - “Looking through” climate-induced shocks may not be viable; monetary authorities may need to act decisively to avoid second-round effects.
- For policy frameworks more broadly:
  - Debate remains unsettled on whether flexible inflation targeting (FIT) is adequate under increasing climate shocks.
  - McKibbin and others (2020) argue FIT may be inadequate and advocate nominal income targeting as it avoids requiring precise understanding of the shock.
  - Cantelmo and others (2022) counter that FIT remains superior in simulations of a small open-economy New Keynesian model with disaster shocks, yielding lower inflation variability and lower consumption-equivalent welfare loss.
- Operational and modeling challenges:
  - Current DSGE models may be ill-suited to model climate shocks because they lack sectoral disaggregation and struggle with “green swan” events and tipping points.
  - Central banks must consider implications for collateral frameworks, asset purchase programs, and potential “green” tilts in operations, balancing monetary policy effectiveness, climate mitigation, risk protection, and operational feasibility (NGFS 2021; Schoenmaker 2019; Papoutsi and others 2021).

### Relationship to existing literature
- Builds on and extends prior work by:
  - Fratzscher and others (2020): natural disasters used as exogenous shocks to evaluate IT in AEs and EMs (their sample: 76 countries).
  - Parker (2018): covers over 220 countries and territories; finds limited effects in AEs but significant and multi-year effects in developing economies.
  - Klomp (2020): analysis of 400 earthquakes for 85 countries (1960–2015) finds disaster effects consistent with supply shocks (lower output, upward pressure on inflation).
  - Cantelmo and others (2022): small open-economy New Keynesian model suggesting FIT is welfare-maximizing in the presence of disaster shocks.
  - McKibbin and others (2020): argue for nominal income targeting in the face of climate-policy-induced shocks.
- Contribution: first analysis to jointly exploit natural disaster events and temperature/precipitation deviations covering a representative set of countries including a large panel of LIDCs.

### Organization of the paper
- Section 2: Related literature.
- Section 3: Data and transformations; stylized facts on natural disasters, temperature, precipitation, and inflation dynamics.
- Section 4: Local projection model and empirical results.
- Section 5: Conclusion and policy implications.
- Appendices cover subnational analysis, additional controls, role of food inflation, impact of floods on food production, alternative measures of monetary autonomy, and alternative measures of temperature and precipitation.

*IMF Working Papers — How Persistent are Climate-Related Price Shocks — International Monetary Fund*

### 3. Data and Stylized Facts

### 3. Data and Stylized Facts

### 3.1 Data Sources and Transformations
- Two datasets capture climate shocks:
  - EM-DAT database (Centre of Research on the Epidemiology of Disasters (CRED), Université Catholique de Louvain) for natural disaster events.
  - High-resolution gridded monthly temperature and precipitation from Climate Research Unit (CRU TS v. 4.03), University of East Anglia.
- EM-DAT disaster inclusion criteria: events where 10 or more deaths are recorded, 100 or more people are affected, resulting in a call for state of emergency and need for international assistance.
- Natural disasters classification:
  - Natural: droughts, earthquakes, floods, storms, wildfires.
  - Technical: industrial and transport accidents (analysis focuses on climate-related natural disasters: droughts, floods, storms).
- Disaster intensity measure (IMF (2003); Becker and Mauro (2006)):
  - Dummy equals 1 when (number of fatalities + 0.3 × number of affected people) > a*, 0 otherwise.
  - Moderate natural disasters: a* = 0.0001 (0.01 percent of the population) as in Fomby, Ikeda, and Loayza (2013).
  - Severe natural disasters: a* = 0.01 (1 percent of the population) as in Becker and Mauro (2006).
- Natural disaster shocks at quarterly frequency:
  - D_{i,t}^k = sum of disaster event dummies of type k occurring in quarter t for country i (combines frequency and intensity).
- Temperature and precipitation construction:
  - Country monthly series from area-weighted average of grid cells within each country, then averaged within each quarter to obtain quarterly series.
  - Climate shocks constructed as country-specific z-scores relative to the past 30 years preceding the quarter:
    - T_{i,t}^{z} = (T_{i,t} - T_{i,t}^{ave}) / σ_{i,t}^T.
    - Precipitation z-scores defined analogously.
  - Long-run averages and volatilities are quarter-specific 30-year statistics.
- CPI and sample construction:
  - Monthly CPI for 183 countries over 1970-2018 from IMF International Financial Statistics (IFS).
  - Monthly CPI converted to quarterly by monthly averages; year-on-year inflation calculated for each country.
  - Countries require at least 10 years of observations.
  - Inflation data restricted to the range ±20 percent to account for hyperinflation episodes.
  - Final merged quarterly datasets cover 1970-2018.
- Country and monetary regime classifications:
  - Country groups per IMF World Economic Outlook classification: advanced economies (AE), emerging markets (EM), low income and developing countries (LIDC).
    - Sample includes 33 AEs, 76 EMs, and 53 LIDCs for country classification.
  - Monetary regime: inflation-targeting (IT) and non-inflation-targeting (non-IT) using Fratzscher and others (2020).
    - Monetary policy categories sample: 28 IT and 73 non-IT.
  - Analysis of monetary policy effects conducted on post-1990 data only (IT first adopted in New Zealand in 1990).

### Stylized Facts on Natural Disasters and Climate Indicators
- Frequency and trends (EM-DAT observations):
  - Overall increase in occurrences of almost all types of natural disasters from 1970-1995 to 1996-2018.
  - Flood events: averaged less than 50 each year between 1970-1995; more than tripled between 1996-2018.
  - In the past two decades (roughly 1999–2018), floods averaged about 150 events per year; storms about 100 occurrences per year; droughts occur less frequently and ranked 7th between 1996 and 2018 but have been rising.
  - The increase in recorded events may be partly driven by improved recording, but this effect is relatively minor.
- Cross-country patterns:
  - Average number of natural disasters increased over time across AEs, EMs, and LIDCs.
  - Largest increase registered for LIDCs: natural disasters occurred once every two years between 1970-1995 and increased to twice every year between 1996-2018.
  - In the most recent two decades, AEs, EMs, and LIDCs had a similar number of natural disasters, with an average of two events per year.
- Frequency statistics (Table 1 highlights):
  - About 31.9 percent of country-quarter observations are associated with at least one natural disaster.
  - By disaster type (portion of at least one event occurring):
    - Floods: 9 percent.
    - Storms: 5.2 percent.
    - Droughts: 1.6 percent.
  - Multiple events in a quarter occur for some countries (e.g., more than one flood or storm in a quarter).
  - Moderate and severe disasters are rarer (e.g., moderate drought: 0.1 percent; almost no countries experienced two severe floods in one quarter).
- Distribution by development status:
  - Larger fraction of AEs affected by storms than EMs and LIDCs.
  - Larger fraction of EMs and LIDCs affected by droughts and floods than AEs.
  - Almost all LIDCs (53 out of 54 in the sample) have been affected by floods.
  - As disaster intensity increases, the number of affected countries in the sample declines.
  - None of the AEs in the sample are affected by moderate or severe drought under the paper’s definitions.
- Summary statistics (Table 2, 1970-2018):
  - Inflation (overall mean): 5.7 percent.
    - AEs mean inflation: 4.5 percent.
    - EMs mean inflation: 6.0 percent.
    - LIDCs mean inflation: 6.3 percent.
    - Non-Inflation Targeting mean inflation: 6.4 percent.
    - Inflation Targeting mean inflation: 3.3 percent.
  - Temperature z-score (overall mean): 0.5.
    - Temperature z-score higher in EMs and LIDCs than in AEs; EMs and LIDCs on average more than 50 percent higher than AEs.
  - Precipitation z-score (overall mean): 0.03.
    - Average precipitation z-scores close to zero; distributions skewed right (more likelihood of positive precipitation shocks).
    - On average precipitation z-scores higher for AEs than EMs and LIDCs.
  - Subsample (1990-2018) statistics for IT/non-IT reported separately in Table 2.
- Distributional patterns (Figure 3 observations):
  - Temperature z-score distributions:
    - Slightly thinner tails for LIDCs than AEs and EMs, consistent with many LIDCs being in tropical regions with more stable temperatures.
    - Distribution of temperature z-scores is indistinguishable across monetary policy regimes (IT vs non-IT).
  - Precipitation z-score distributions:
    - AEs tend to experience larger deviations from historical averages.
    - Precipitation z-score distributions similar across monetary policy regimes.

### Data Usage and Comparability Notes
- Natural disaster analysis uses a quarter-level count of events combining frequency and intensity (equation (2) specification).
- Temperature and precipitation shocks are separate analyses due to differing shock specifications (equation (1) vs equation (3)).
- The baseline empirical models do not include control variables because of limited quarterly data availability; robustness checks with controls (nighttime lights, exchange rate, food weights) are considered in the appendix.
- Country coverage and minimum observation requirements ensure dynamic impact estimation is feasible (countries with at least 10 years of observations included).

*Source: wpiea2022207-print-pdf - 3. Data and Stylized Facts*

### 4.2 Impact of Natural Disasters on Inflation

### 4.2 Impact of Natural Disasters on Inflation

### Heterogeneity across different natural disasters
- Droughts:
  - An average drought raises inflation by 0.2 percentage point upon impact, with a statistically significant response at the 95 percent level.
  - The impact lingers at a similar level for the next four quarters but becomes statistically insignificant two quarters after the shock.
  - A moderate or severe drought raises inflation by 3 percentage points within four quarters.
  - The impact of a moderate drought peaks in the fourth quarter; the impact of a severe drought peaks in the first quarter.
  - Overall, the impact of drought on inflation is temporary, lasting less than a year.
- Floods:
  - Average floods exert a deflationary reaction, but the impact is statistically insignificant even at 68 percent level.
  - Severe floods trigger a strong and statistically significant response (at 95 percent level), reaching 2 percentage points a year after the shock.
  - The negative response to floods is consistent with demand shocks (risk-aversion) dominating supply shocks in many cases.
- Storms:
  - Storms are initially inflationary; effects are short lived.
  - Significant for about two and three quarters for moderate and severe events, respectively.
  - Subsequently the response becomes negative and remains depressed for more than two years.
  - Both magnitude and duration, and significance level, increase with storm intensity.
- Mechanisms and interpretation:
  - Differential impacts reflect relative changes in aggregate supply and aggregate demand: if aggregate supply is reduced by more than aggregate demand, inflationary effects follow; demand shocks operate through heightened risk aversion, reducing consumption and investment and dampening inflation.
  - For floods, subdued demand often dominates supply effects, producing deflationary outcomes empirically.
- Role of food:
  - Food’s weight in consumption baskets (especially in EMs and LIDCs) can dominate headline inflation responses.
  - Appendix C: countries with a larger share of consumption basket in food face higher inflationary pressures following droughts, floods, and storms.
  - Appendix D: production of major food items tends to increase following floods, which could lead to decreases in food prices and thus lower headline inflation after floods.

### Temporal and spatial factors
- Temporal:
  - Floods tend to be short-lived and often occur at the beginning or middle of the rainy season; farmers may harvest replanted crops.
  - Droughts tend to last longer.
- Spatial:
  - Floods usually cover smaller land areas (near major rivers or lowlands); droughts generally cover larger areas.
  - Flood damage is more spatially concentrated; increased precipitation outside immediate flood zones can increase output and induce negative inflationary shocks.

### Heterogeneity across different country types (AEs, EMs, LIDCs)
- Advanced Economies (AEs):
  - All events (droughts, floods, storms) are inflationary in AEs on average.
  - Drought: inflation rises by 2 percentage points in the second quarter, stays elevated about three quarters, then declines; estimates not different from zero even at 68 percent level.
  - Flood: inflation gradually increases by 0.2 percentage point within two years, statistically significant at 68 percent level.
  - Storm: raises inflation above 0.2 percentage point within one year, statistically significant at 95 percent level.
  - Severe floods in AEs: strong deflationary impact in the first six quarters, decreasing inflation to -5 percentage points within a year, then becoming inflationary and significant the next year.
  - Moderate floods in AEs: impact remains inflationary.
  - Moderate and severe storms in AEs: have a deflationary impact (contrasting with average storms).
- Emerging Markets (EMs):
  - Droughts tend to be inflationary.
  - Floods and storms tend to be deflationary.
  - Magnitude increases with intensity: an average flood reduces inflation by less than 0.5 percentage point within a year; a moderate flood reduces inflation by 1 percentage point; a severe flood reduces inflation by almost 4 percentage points.
  - An average drought is inflationary for EMs; severe droughts could exert deflationary pressures after one year.
- Low-Income Developing Countries (LIDCs):
  - Similar pattern to EMs: droughts tend to be inflationary; floods and storms tend to be deflationary.
  - Intensity amplifies magnitude of impacts.
- Implication:
  - As intensity of natural disaster increases, both magnitude and duration of inflationary/deflationary impacts tend to be larger and more persistent.

### Heterogeneity across different monetary regimes (Inflation Targeting (IT) vs non-IT)
- General:
  - Countries classified as IT versus non-IT show different inflation responses to natural disasters.
  - Evidence suggests monetary policy contains inflationary pressures after natural disaster shocks in both AEs and EMs, while it tends to accommodate deflationary impacts.
- Advanced Economies (IT vs non-IT):
  - IT countries: initial inflation reaction of less than 0.5 percentage point for droughts; point estimates of response are negative from the second quarter onward.
  - Non-IT countries: inflation rises and remains elevated at about 2 percentage points for about two years before decreasing to zero.
  - Floods and storms: slightly deflationary in IT countries among AEs; inflationary in non-IT countries.
  - Interpretation: monetary authorities in IT countries have managed to contain inflationary pressures from natural disasters.
- Emerging Markets (IT vs non-IT):
  - IT countries: inflation is anchored around zero for droughts and floods; storms’ impact increases inflation in the second year and remains by 0.1 percentage point.
  - Non-IT countries: drought generates a 3 percentage points jump in inflation; floods increase inflation gradually reaching 1 percentage point in the second year; storms reduce inflation by 1 percentage point in the second year.
- Appendix E: additional evidence that countries with stronger monetary policy frameworks are better insulated against inflationary pressures following natural disasters.

### Food inflation
- Aggregate findings:
  - Droughts cause sharp jumps in food inflation: food inflation jumps sharply to over 5 percentage points upon a drought shock, then rises steadily to a maximum close to 10 percentage points in the third quarter after impact.
  - Food inflation stays elevated for the first year after a drought, then reverses into negative territory for about five quarters.
  - Floods: effect on food inflation is mostly deflationary in the first year (consistent with headline inflation).
  - Storms: food prices rise for four quarters following storms, but the effect is much smaller than for droughts.
- Subnational and item-level findings (Appendix A):
  - Droughts have a stronger impact on staples inflation in EMs than in LIDCs.
  - Droughts show initial inflationary dynamics for some food items (staples, meat, fish); initial deflationary responses are evident for other items (eggs, milk).
  - Subnational analysis reveals large heterogeneity: local inflation dynamics for individual food items can differ markedly from national-level patterns and depend on food category, disaster type, and monetary regime.

*IMF Working Papers — Excerpt from "How Persistent are Climate-Related Price Shocks" (Section 4.2).*

### 5. Conclusion

### 5. Conclusion

### Main empirical findings on climate shocks and inflation
- The frequency and intensity of climate shocks have increased, consistent with climate change.
- Not all climate shocks have similar impacts on inflation:
  - Droughts—although much less frequent than floods—tend to increase inflation.
  - Floods are more likely to dampen inflation.
  - Storms generally exert inflationary pressures and can be persistent.
- The impact of droughts on inflation is largely transmitted via effects on food prices.
- Historical overall impacts on inflation have generally been muted, but given the likely increase in the frequency and amplitude of climate shocks, the trend is for the impact to increase over time.
- Monetary policy regime matters materially for inflation dynamics:
  - Inflation-targeting countries (and more generally countries with stronger monetary policy frameworks) tend to withstand the inflationary impact better than others, consistent with Fratzscher and others (2020).
  - Evidence using an alternative index of monetary autonomy (IAPOC, 0 to 1) provides weak supporting evidence that countries with stronger monetary policy frameworks are better insulated against natural disasters’ impact on inflation.

### Implications for monetary policy frameworks
- Commonly used monetary policy frameworks and underlying models are generally ill-suited to address supply shocks, especially if such shocks are increasing in frequency and amplitude.
- It would not be appropriate to “look through” supply shocks if they become part of the persistent landscape.
- Tightening monetary policy in response to climate-driven supply shocks could lead to output loss and make monetary policy less potent, especially if the equilibrium real interest rate (r*) declines with climate change (as argued by Bylund and Jonson (2020)).
- The debate on alternative frameworks is unsettled:
  - McKibbin and others (2020) propose nominal income targeting as less demanding for policymakers because it does not require precise understanding of the nature of climate shocks.
  - Cantelmo and others (2022) contend that the FIT paradigm is still superior to alternatives they simulated.
  - More research is needed to decide on appropriate frameworks and to account for heterogeneous impacts by type, duration, and intensity of climate shocks.

### Subnational analysis: natural disasters and food inflation (Appendix A)
- Data and coverage:
  - Geocoded Disasters (GDIS) extends EM-DAT with subnational locations for natural disasters between 1960 and 2018.
  - World Food Programme (WFP) Vulnerability Analysis and Mapping provides food price data since 1990 for 98 countries, 637 first administrative regions, and 2,475 markets; the subnational sample is restricted to 1990–2018 and regions with at least 5 years of data for each food category.
  - Food categories analyzed include staples, pulses, vegetables, fruit, meat and fish, eggs and milk, sugar, and oil.
- Key results for staples (subnational, monthly):
  - Upon impact of a drought, staples inflation gradually increases for 9 months before subsiding; the impact in the second year following the drought is slightly deflationary.
  - Floods have a deflationary impact 6-12 months after occurrence.
  - Storms are inflationary and seem persistent.
  - Magnitudes at the subnational level tend to be larger and more precisely estimated than national quarterly estimates.
  - Heterogeneity by income status:
    - Droughts drive staples inflation in EMs to well above 10 percentage points after a muted reaction of about five months; their impact on staples inflation in LIDCs is limited.
    - Floods’ deflationary impact in the 6-9 months is more visible for LIDCs than for EMs.
    - Storms exert inflationary pressures on EMs in the first year, while such pressures manifest in the second year for LIDCs.
  - Heterogeneity by monetary regime:
    - The inflationary impact of droughts and storms on staples is stronger in IT-countries, raising staples’ inflation by 10 percentage points in the first year of impact—more than double that for non-IT countries.
- Inflation response across other food categories:
  - Droughts have an inflationary impact on meat and fish in the first 12 months, but a deflationary impact on eggs and milk.
  - Floods’ deflationary impact is concentrated in sugar.
  - Storms have a sizeable deflationary impact on eggs and milk.
- Conclusion from subnational analysis: substantial heterogeneity exists across food categories, disaster types, regions, and monetary regimes; local inflation dynamics can differ markedly from national dynamics.

### Robustness and additional controls (Appendix B)
- Extended baseline model includes additional controls: exchange rate growth (International Financial Statistics), average night lights (DMSP-OLS and VIIRS series), and food weight in the CPI basket (IMF data).
- With additional controls, results are broadly similar to baseline Figures 5–7; sample size declines and some disaggregated results are missing.
- Notable differences for LIDCs: natural disaster shocks induce inflationary responses in almost all cases, with the marked exception of storms for average moderate cases; severe storms are inflationary for just about more than a year.

### The role of food share in CPI (Appendix C)
- Median weight of food in CPI consumption basket:
  - Advanced economies: around 15%
  - Emerging markets: 25%
  - Low-income countries: 45%
- Higher share of food in the consumption basket leads to more inflationary pressures following natural disaster shocks, especially in the short term.
- With growing food share, droughts exert more inflationary pressures than floods and storms.

### Floods and food production (Appendix D)
- Using FAOSTAT annual production data and aggregating disaster data to annual frequency:
  - For rice, maize, and meat, for all floods on average food production increases in ensuing years after impact.
  - Moderate floods tend to be beneficial for food production.
  - Severe floods’ impact tends to be negative in the initial years.

### Alternative measure of monetary autonomy (Appendix E)
- IAPOC (Unsal and others (2022)) ranges from 0 to 1 and captures independence, accountability, strategy, and communications; expanded from annual to quarterly by repeating yearly values.
- Improvement in IAPOC does not imply a uniform directional change in inflation responses:
  - Stronger monetary policy frameworks are not associated with systematically lower or higher inflationary impact of drought.
  - Stronger frameworks seem associated with lower inflationary impact of floods and higher inflationary impact of storms.
- Comparing countries above vs. below median IAPOC: responses are mostly not significant; confidence intervals are wider for countries with weaker frameworks.
- Taken together, evidence is suggestive but weak that stronger monetary policy frameworks provide insulation against natural disasters’ inflationary impacts.

*Source: 5. Conclusion and accompanying appendices from wpiea2022207-print-pdf - 5. Conclusion*

### Appendix F. Alternative Measure of Temperature and

### Appendix F. Alternative Measure of Temperature and Precipitation

### Methodology and data
- Considered measure: the 3-month Standardized Precipitation Evapotranspiration Index (SPEI).
- Data source: The Global SPEI database.
- SPEI construction: based on monthly precipitation and potential evapotranspiration from the Climatic Research Unit of the University of East Anglia.
- Rationale: This index combines both temperature and precipitation into a single measure to capture persistent climate shocks; for example, droughts are brought about by higher temperature and lower precipitation, making the SPEI a suitable aggregate measure to consider.

### Key findings
- The impact of positive and negative SPEI on inflation is presented by country income status.
- Results are qualitatively similar to those for precipitation shocks in Figure 12.
- Negative SPEI shocks are inflationary for:
  - Emerging markets (EMs)
  - Low-income and developing countries (LIDCs)
- Figure panels summarize impacts by group:
  - (a) AEs
  - (b) EMs
  - (c) LIDCs

### Figure
- Figure F1. Impact of Positive and Negative SPEI on Inflation
- Figure source: The Global SPEI database, International Financial Statistics, and authors’ calculations.

*How Persistent are Climate-Related Price Shocks? Implications for Monetary Policy — Appendix F. Alternative Measure of Temperature and Precipitation*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022207-print-pdf.pdf_