## wpiea2024058-print-pdf

## Source details

**Canonical URL:** [wpiea2024058-print-pdf](https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024058-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2024/english/wpiea2024058-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2024/english/wpiea2024058-print-pdf.pdf.json)

---

### Overview and purpose
- Provides supplemental material (figures, tables, and summary statistics) to support the paper "Climate Variability and Worldwide Migration: Empirical Evidence and Projections."
- Aims to provide estimates of migration flows between countries in response to changes in temperature, precipitation, droughts, and excess precipitation under several socio-economic and climate scenarios.
- Results intended to support policymakers regulating migration and assessing societal and macroeconomic impacts of climate-induced population movements.

### Major empirical findings and projections
- Core empirical estimates:
  - "A one percent decadal increase in temperature increases decadal average bilateral migration rates by 1.2 percent."
  - Meta-analysis and model coefficients report ln(temperature) estimates including 1.168*** (standard error (0.283)) and other similar significant estimates across specifications.
- Projected global decadal emigration flows (ranges across RCPs and SSPs):
  - 2030 decade: "73 and 91 million".
  - 2040 decade: "83 and 102 million".
  - 2050 decade: "88 and 121 million".
  - 2060 decade: "87 and 133 million".
- Regional/worst-case and comparative estimates:
  - "5.3 additional migrants every one thousand people will migrate annually from sub-Saharan Africa by the end of the 21st century."
  - Translating population change, predictions imply "yearly climate migration of up to 18.5 million inhabitants in the worst climate scenario."
  - Missirian and Schlenker (2017): asylum applications to Europe increase by "28 percent" under RCP 4.5 and "188 percent" under RCP 8.5, translating into "98,000" and "660,000" additional asylum applications per year, respectively.
  - Groundswell (Rigaud et al. 2018): up to "143 million" people could migrate domestically between "2020 and 2050" in Sub-Saharan Africa, South Asia, and Latin America ("4.8 million additional migrants per year, on average"); expanded regions yield "216 million" ("7.2 million additional migrants per year, on average").
  - Burzyński et al. (2022): "37, 57, and 94 million" total international climate migrants over the 21st century under RCPs "4.5, 7.0, and 8.5," respectively; under RCP 7.0 easing mobility barriers increases migrants from "57 to 77 million."
- Decomposition takeaway:
  - "The most important driver of migration is population growth, while changes in decadal average climatic conditions alone are responsible for a small fraction of the projected migration flows."
  - Climate-only contribution: in most cases, climate change increases outflows by much less than 5 million per year; when population is held constant, climate migrants are approximately equal to the average number of migrants from 1970 to 2010.

### Methodology and data
- Empirical framework:
  - Bilateral gravity equation for bilateral emigration rates estimated with Poisson-pseudo maximum likelihood (PPML).
  - Specification includes origin fixed effects (표표_i), destination-time fixed effects (d_j,t), and bilateral dummies (ψ_ij). Standard errors clustered at the dyadic level (ij).
  - Identification from random variation in origin countries' decadal averages of weather and migration from long-term means; parsimonious specification avoids controlling for socio-economic variables endogenous to climate.
- Sample and temporal coverage:
  - Panel of "100 origin countries" and "166 destination countries" over five decades (decades beginning "1960, 1970, 1980, 1990, 2000").
  - Analysis covers years from "1960 to 2010", grouped in decades indexed with their first year (t = "1960" up to t = "2000").
  - OECD origin countries excluded, except "Mexico and Chile" (robustness check drops them).
- Migration data:
  - Primary source: Abel (2018) bilateral migration flows computed via demographic accounting using census-based foreign-born stocks covering "1960 up to 2010".
  - Input stock data: Global Bilateral Migration Database (Ozden et al. 2011) for "1960 to 2000" and United Nations Population Division (2015a) thereafter; demographic inputs from United Nations Population Division (2015b).
  - Emigration rates computed by dividing bilateral flows by origin country population (United Nations Population Division (2015b)).
- Climate data and indicators:
  - Historical climate: NASA NOAH GLDAS v2 at "1°×1°" spatial resolution and 3-hourly intervals; aggregated to country-year using 2000 population weights (CIESIN 2018); then averaged by decade.
  - Drought indicator: 12-month Standardized Precipitation Index (SPI); count months SPI < "-1.5" in grid cells with crops (FAO and IIASA 2022); population-weighted decadal averages.
  - Flood-like indicator: "90th percentile" of annual distribution of daily precipitation in each grid cell; population-weighted decadal averages.
  - Future climate: five GCMs in NEX-GDDP (CMIP5) ensemble mean for RCP4.5 and RCP8.5.
- Population scenarios:
  - Gridded population projections for the SSPs (O'Neill et al. 2017) used for weighting and predicting migration.
  - Scenario pairing: RCP4.5 matched with SSP1, SSP3, and SSP5; RCP8.5 matched with SSP5 only.

### In-sample empirical results and robustness
- Specification and robustness checks (Table 1, Equation (1), six columns):
  - Temperature effects consistent in magnitude and significance across specifications.
  - Point estimates:
    - If temperature increases by 1 percent, bilateral migration rates increase by 1.2 percent over a decade (specifications 1–5).
    - Specification (6) indicates a slightly larger increase equal to 1.5 percent.
  - Differences in temperature-effect estimates are never statistically significant across the six specifications.
- Precipitation, droughts, and floods:
  - Average precipitation coefficient not statistically significant; excluding precipitation leaves results unchanged.
  - Drought and flood coefficients not statistically significant in full sample; using only Global Bilateral Migration Database (ends in 2000) (Column 6) yields statistically significant coefficients indicating increases in drought and flood conditions reduce migration rates.
  - Quantified effects in Column 6:
    - A one standard deviation increase in average drought (corresponds to 0.3 months with SPI < -1.5) reduces emigration rates by 7.7 percent.
    - If rain increases by one millimetre during the rainiest days of the year (approximately a 10 percent shock compared to the sample mean), emigration rates decrease by 18.7 percent.
- Nonlinearity checks:
  - Column (5) rejects a nonlinear ln(temp)^2 effect; first-difference model (Column (6) OLS) still finds positive migration response to warming and decreases with floods and droughts.
- Data-construction and robustness:
  - Migration stocks in the Global Bilateral Migration Database were harmonized to be homogeneous between 1960 and 2000 (definitions, borders, aggregation); where census data missing, imputations were used.
  - Robustness checks: (i) using only Global Bilateral Migration Database 1960–2000; (ii) computing flows via Abel’s demographic accounting (main) versus difference-in-stocks (robustness).

### Projections, scenarios, and sensitivity analyses
- Simulations and projection methodology:
  - Parameters estimated with Equation (1); climate variables computed using GCM scenarios; bootstrap confidence intervals generated using 1,000 repetitions.
  - Conversion from projected emigration rates to migration flows uses M̂_ij,t = ṽ̂_ij,t * N_i,t.
  - Additional scenarios: globally uniform temperature increases of +1°C, +2°C, and +3°C while other climatic indicators remain as observed (to isolate temperature effect).
  - Bilateral approach enables predicting outflows and corresponding inflows across destinations.
- Projection sensitivities and caveats:
  - Exercise is ceteris paribus: assumes only climate and population change; socio-economic, technological, and policy drivers held constant.
  - Projecting to 2060 and under RCP 8.5 extends climate variables beyond historical variance used for estimation; bias direction ambiguous (may understate nonlinear/adaptive responses).
  - Uniform warming scenario: total migration increases with warming but projections are never significantly different from historical average flow at the 95 percent level.
  - Climate-only contribution: in most cases, climate increases outflows by much less than 5 million per year.

### Interpretation and policy implications
- Migration as adaptation and macroeconomic effect:
  - Migration can generate large positive welfare effects by enabling escape from severe climate impacts, facilitating labor reallocation, and limiting costs of sea-level rise.
  - Large emigration from small countries may jeopardize macroeconomic stability; immigration into aging countries may ease long-term macroeconomic constraints.
- Policy recommendations and considerations:
  - Policymakers need origin-destination-size estimates to plan migration management and anticipate macroeconomic and societal challenges.
  - Empirically grounded projections that account for socio-economic context, costs, and constraints provide more credible estimates than approaches assuming all at-risk populations will migrate.
  - Restrictive border policies do not reduce vulnerability to climate change and can limit people’s ability to adapt.
  - Policies to facilitate temporary migration in the case of extreme events would be beneficial.
  - Destination countries concerned about rising migration can pursue policies that foster slower population growth and higher economic development in origin countries (examples: investing in women’s empowerment).
- Social risks and uncertainties:
  - Mass migration may create competition for scarce land and resources and lead to cultural conflict; societal impacts are hard to quantify and potentially very large.
  - Out-of-sample and long-term uncertainties remain substantial; population growth is the dominant projected driver, while climate-average changes explain a small share.

### Annex I — Key tables and summary statistics (Table AI.1)
- Emigration Rate (*1000) – bilateral: Obs. 74,437; Mean 0.257; Sd 3.171; Min 0; Max 215.845
- Emigration Flows – bilateral: Obs. 74,437; Mean 2,408; Sd 40,200; Min 0; Max 5,077,120
- Temperature (Degree C) – unilateral: Obs. 74,437; Mean 23.484; Sd 5.053; Min 0.082; Max 32.347
- Precipitation (annual mm) - unilateral: Obs. 74,437; Mean 1185; Sd 743; Min 28; Max 3377
- Droughts (weighted number of months) - unilateral: Obs. 74,437; Mean 0.225; Sd 0.274; Min 0; Max 1.000
- Flood (mm per day) - unilateral: Obs. 74,437; Mean 9.113; Sd 6.555; Min 0; Max 27.011

*Source: Annex I. Additional Tables, IMF Working Paper "Climate Variability and Worldwide Migration: Empirical Evidence and Projections."*

### Annex I. Additional Tables .............................................................................................

### Annex I. Additional Tables

### Overview and purpose
- Provides supplemental material (figures, tables, and summary statistics) to support the paper "Climate Variability and Worldwide Migration: Empirical Evidence and Projections."
- Aims to provide estimates of migration flows between countries in response to changes in temperature, precipitation, droughts, and excess precipitation under several socio-economic and climate scenarios.
- Results are intended to support policymakers regulating migration and assessing societal and macroeconomic impacts of climate-induced population movements.

### Key background and conceptual points
- Historical and archaeological evidence shows migration in response to environmental change; examples cited include societal collapse linked to abrupt climate changes and urban population responses to freshwater availability in highland Mexico civilizations.
- Present-day relevance of historical evidence is limited because:
  - Mobility and communication are easier now.
  - Borders are more tightly controlled.
  - Technological progress can smooth climate shocks.
  - Aid provides temporary relief from food shortages.
- Migration decisions are influenced by economic, social, political, and demographic contexts; not all people exposed to climate hazards will migrate because costs, constraints, and other factors matter.

### Welfare and macroeconomic implications of migration
- Migration can generate large positive welfare effects by:
  - Allowing individuals to escape severe climate impacts.
  - Facilitating labor reallocation to more productive areas (Kahn 2010).
  - Gradual migration from coastal zones limiting the cost of sea-level rise (Diaz 2016).
- Negative consequences occur when migration is restricted or when people are trapped in high-risk areas due to policy or resource constraints (Benveniste, Oppenheimer, and Fleurbaey 2020; 2022).
- Mass migration may create competition for scarce land and resources and can lead to cultural conflict; these impacts are hard to quantify but potentially very large.
- Macroeconomic effects include risks to GDP growth potential:
  - Large emigration from small countries may jeopardize macroeconomic stability.
  - Immigration into aging countries may ease long-term macroeconomic constraints.
- Effective migration management requires information on origin, destination, and size of potential flows.

### Existing literature and methodological approaches
- Early projections that assumed all people at risk will migrate produced very large estimates (e.g., claims up to 200 million per year by 2050) that are not endorsed by the scientific community.
- More recent empirical approaches use micro-level data to estimate relationships between migration and climatic variables, accounting for individual preferences and constraints. These use econometric estimates to predict changes in migration probability under climate scenarios, holding other factors constant.
  - Such micro-level studies typically focus on single countries and capture local specificities but cannot be extrapolated globally.
- Examples of micro-level estimates preserved verbatim:
  - Jessoe, Manning, and Taylor (2017): predict that a medium emissions scenario will increase the probability of migration to urban areas by as much as 1.4 percent and to the USA by as much as 0.2 percent in 2075. This translates into 230,000 additional migrants to urban centers and 40,000 additional migrants to the US each year.
  - Cattaneo and Massetti (2019): forecast 6.3 and 3.6 million additional migrants within Nigeria in the period 2071–2100 considering RCP 8.5 and RCP 4.5 climate scenarios, respectively.
- Panel-data studies extend analysis across many countries; example referenced but truncated in the source text (Marchiori, Maystadt, and Schumacher 2012).

### Contents listed in Annex I (figures, tables, and annex tables)
- Figures included in the paper (titles preserved exactly as listed):
  - 1. Change of Climate Variables with Respect to the Period 1960-2010
  - 2. Comparison of Average Decadal Historical and Future Temperature
  - 3. Global Decadal Outflows of Climate Migrants (Uniform Scenario)
  - 4. Predicted Global Decadal Outflows of Climate Migrants. Historical Comparison
  - 5. Population Projections
  - 6. Global Decadal Outflows of Climate Migrants. Decomposition for Climate and Population
  - 7. Percentage Change in Outflows Compared to Historical Outflows - Year 2030
  - 8. Bilateral Flows by Continent. SSP3, RCP 4.5 - Year 2030
- Tables included in the paper (titles preserved exactly as listed):
  - 1. Historical Estimations
  - 2. Historical Estimations, Robustness Checks
- Annex tables:
  - AI.1. Summary Statistics

### Implications for policy and research (from source framing)
- Policymakers need origin-destination-size estimates to plan migration management and anticipate macroeconomic and societal challenges.
- Empirically grounded projections that account for socio-economic context, costs, and constraints provide more credible estimates than approaches that assume all at-risk populations will migrate.
- Further global-scale projections require methods that balance micro-level credibility with geographic scope.

*Source: Annex I. Additional Tables, IMF Working Paper "Climate Variability and Worldwide Migration: Empirical Evidence and Projections."*

### 5.3 additional migrants every one thousand people will migrate annually from sub-Saharan Africa by the end of

### wpiea2024058-print-pdf - 5.3 additional migrants every one thousand people will migrate annually from sub-Saharan Africa by the end of

### Major empirical and literature findings
- Climate-induced migration estimates from panel and cross-sectional studies:
  - "5.3 additional migrants every one thousand people will migrate annually from sub-Saharan Africa by the end of the 21st century" in response to climate change.
  - Translating population change, predictions imply "yearly climate migration of up to 18.5 million inhabitants in the worst climate scenario."
  - Missirian and Schlenker (2017) forecast an increase in asylum applications to Europe equal to "28 percent" under RCP 4.5 and "188 percent" under RCP 8.5, translating into "98,000" and "660,000" additional asylum applications per year, respectively.
- Groundswell / World Bank and related cumulative and annual estimates:
  - Groundswell Report (Rigaud et al. 2018) predicts up to "143 million" people could migrate domestically between "2020 and 2050" to escape slow-onset climate impacts in Sub-Saharan Africa, South Asia, and Latin America ("4.8 million additional migrants per year, on average").
  - Expanding to include East Asia and the Pacific, North Africa, and Eastern Europe and Central Asia raises the cumulative number to "216 million" ("7.2 million additional migrants per year, on average").
- Dynamic/global modelling results:
  - Burzyński et al. (2022) predict "37, 57, and 94 million" total international climate migrants over the 21st century under RCPs "4.5, 7.0, and 8.5," respectively (changes relative to a no-climate-change 2010 climate baseline).
  - Under RCP 7.0, easing mobility barriers would increase migrants from "57 to 77 million."
- Key empirical result from this paper:
  - "A one percent decadal increase in temperature increases decadal average bilateral migration rates by 1.2 percent."
  - Emigration flows projected across decades and scenarios:
    - "73 and 91 million" during the 2030 decade (average decadal emigration flows).
    - "83 and 102 million" for the 2040 decade.
    - "88 and 121 million" for the 2050 decade.
    - "87 and 133 million" for the 2060 decade.
  - Decomposition result: "the most important driver of migration is population growth, while changes in decadal average climatic conditions alone are responsible for a small fraction of the projected migration flows."

### Methodology and data summary
- Empirical framework:
  - Bilateral gravity equation for bilateral emigration rates estimated with Poisson-pseudo maximum likelihood.
  - Specification includes origin fixed effects (표표_i), destination-time fixed effects (d_j,t), and bilateral dummies (ψ_ij). Identification comes from random variation in origin countries' decadal averages of weather and migration from long-term means.
  - Standard errors clustered at the dyadic level (ij).
  - Rationale: parsimonious specification avoids controlling for socio-economic variables that are endogenous to climate.
- Sample and temporal coverage:
  - Panel of "100 origin countries" and "166 destination countries" over five decades (decades beginning "1960, 1970, 1980, 1990, 2000").
  - Analysis covers years from "1960 to 2010", grouped in decades indexed with their first year (t = "1960" up to t = "2000").
  - OECD origin countries excluded, except "Mexico and Chile" (robustness check drops them).
- Migration data:
  - Primary source: Abel (2018): bilateral migration flows computed via demographic accounting using census-based foreign-born stocks and demographic adjustments, covering almost all origin-destination pairs from "1960 up to 2010".
  - Input stock data: Global Bilateral Migration Database (Ozden et al. 2011) for "1960 to 2000" and United Nations Population Division (2015a) for subsequent decade; demographic inputs from United Nations Population Division (2015b).
  - Emigration rates computed by dividing bilateral flows by origin country population (United Nations Population Division (2015b)).
- Climate data and indicators:
  - Historical climate: NASA NOAH Global Land Data Assimilation System (GLDAS v2) at "1°×1°" spatial resolution and 3-hourly intervals; aggregated to country-year using 2000 population weights from NASA's Socioeconomic Data Applications Center Gridded Population of the World (CIESIN 2018); then averaged by decade.
  - Drought indicator: 12-month Standardized Precipitation Index (SPI); count months SPI < "-1.5" in grid cells with crops (using Global Agro-Ecological Zones database (FAO and IIASA 2022)); aggregated to country-level using population weights and decadal averages.
  - Flood-like indicator: "90th percentile" of annual distribution of daily precipitation in each grid cell; aggregated to country-year using population weights and decadal averages.
  - Future climate scenarios from five GCMs in NEX-GDDP (CMIP5) ensemble mean for RCP4.5 and RCP8.5.
- Population scenarios:
  - Gridded population projections for the SSPs (O'Neill et al. 2017) used for weighting and predicting migration.
  - Scenario pairing: RCP4.5 matched with SSP1, SSP3, and SSP5; RCP8.5 matched with SSP5 only.

### Analysis insights and interpretation
- On short-term vs long-term inference:
  - Panel studies extrapolate effects of "relatively small, short-term, and unpredictable weather shocks" to "relatively large, long-term, and at least in part predictable climate change"; equivalence holds only "under very restrictive assumptions" (Hsiang 2016); direction of bias is difficult to predict (Ionesco, Mokhnacheva, and Gemenne, 2016).
  - Cross-sectional adaptation-based approaches (e.g., Cattaneo and Massetti 2019) assume cold places will adapt like currently warm places, but may be biased by omitted variables and do not capture transition costs.
- International versus internal migration:
  - Most climate-induced moves expected to be domestic because international moves are more costly and tightly controlled.
  - Peri and Sasahara (2019) predict divergent internal migration responses by income group:
    - Poor countries: internal migration rates decline from "7.1 percent" in 2000 to "5.0 percent" (pessimistic) and "5.5 percent" (optimistic) by 2080-2100.
    - Upper-middle-income countries: internal migration rates increase from "7.0 percent" in 2000 to "7.8 percent" (pessimistic) or remain constant.
- Role of climate extremes:
  - Literature and this paper emphasize importance of extreme temperature, drought, and flood events as drivers of economic outcomes and potential migration (Akyapi, Bellon and Massetti 2022); droughts and floods historically responsible for the largest number of deaths.
  - This analysis incorporates drought and extreme precipitation (flood-like) measures in addition to average temperature and precipitation.

### Projections and comparative scenario outcomes
- This paper’s projected global decadal emigration flows by decade (ranges across RCPs and SSPs, exact scenario breakdowns provided in full paper):
  - 2030 decade: "73 and 91 million".
  - 2040 decade: "83 and 102 million".
  - 2050 decade: "88 and 121 million".
  - 2060 decade: "87 and 133 million".
- Comparison to other major studies:
  - Worst-case regional estimate: up to "18.5 million" yearly climate migrants in sub-Saharan Africa under worst climate scenario (from initial statement).
  - Groundswell: cumulative "143 million" (2020–2050) = "4.8 million per year"; expanded regions "216 million" = "7.2 million per year".
  - Burzyński et al. (2022): international climate migrants over 21st century "37, 57, and 94 million" under RCPs "4.5, 7.0, and 8.5"; mobility easing increases "57 to 77 million" under RCP 7.0.
- Decomposition takeaway:
  - Population growth is the dominant driver of the projected increase in migration flows; decadal average climatic changes account for only a small fraction of projected flows.

### Modelling caveats and robustness
- Identification relies on deviation of origin countries' decadal climate averages from their long-term means; rich fixed effects absorb many confounders.
- Use of decadal averages (rather than annual) intended to better capture medium-term climate change effects and adaptation.
- Climate extremes constructed from gridded precipitation rather than disaster-reported datasets to reduce endogeneity from reported impacts.
- SPI chosen over SPEI because potential evapotranspiration projections needed for SPEI are not straightforward for future projection construction.
- Sample excludes OECD origin countries (except Mexico and Chile) because climate likely plays a minor role in migration from richer countries; robustness checks drop Mexico and Chile.

*Italic source: IMF Working Paper — Climate Variability and Worldwide Migration: Empirical Evidence and Projections (excerpts from wpiea2024058-print-pdf).*

### 2000. Migration stocks in the Global Bilateral Migration Database were constructed to be homogeneous

### 2000. Migration stocks in the Global Bilateral Migration Database were constructed to be homogeneous

### Data construction and robustness checks
- Migration stocks in the Global Bilateral Migration Database were harmonized to be homogeneous concerning:
  - the definition of migrants,
  - any changes that occurred in the borders and identity of the countries,
  - the aggregation of the origin countries recorded in censuses.
- Where census data were missing, different forms of imputation were used.
- The harmonization makes the data consistent between 1960 and 2000, but may complicate comparisons with the other dataset that starts in 2000.
- Robustness tests implemented:
  - Using only the Global Bilateral Migration Database from 1960 to 2000 to test effects of differences in standardization rules.
  - Computing flows from stocks via two methods:
    - Abel’s migration flows predicted using a demographic accounting method (main estimation).
    - Flows computed as a difference between two subsequent decades of the stock using the Global Bilateral Migration Database (robustness check).

### Simulations and projection methodology
- Emigration rate predictions:
  - Parameters estimated with Equation (1).
  - Climate variables computed using scenarios from General Circulation Models.
  - Bootstrap confidence intervals generated using 1,000 repetitions.
- Conversion from projected emigration rates to migration flows:
  - 푀푀
    푖푖푗푗푖푖
    =푦푦�
    푖푖푗푗푖푖
    ∗ 푁푁
    푖푖푖푖,푖푖
    .
- Additional scenarios:
  - Projected impacts of a globally uniform temperature increase equal +1°C, +2°C, and3°C, while all other climatic indicators remain as observed in the past (hypothetical uniform warming scenario to isolate temperature effect).
- Use of bilateral data:
  - Bilateral (origin-destination) approach allows predicting both outflows and corresponding inflows across destination countries, enabling fine-grained predictions of migration patterns induced by climate change.

### In-sample empirical results (gravity model estimations)
- Specification overview (Table 1, Equation (1)):
  - Column (1): baseline specification, full dataset.
  - Column (2): excludes average precipitation from climatic controls.
  - Column (3): excludes controls for droughts and floods.
  - Column (4): replaces origin-destination dummies with bilateral controls (bilateral distance, contiguity, common colonial history, common language).
  - Column (5): excludes Chile and Mexico from origin sample (sample then includes only non-OECD countries).
  - Column (6): runs specification (1) but excludes years after 2000 to check effect of changing stock migration data source for years after 2000.
- Temperature effects:
  - Across all six specifications, estimated parameters for temperature are consistent in magnitude and significance.
  - Point estimates:
    - If temperature increases by 1 percent, bilateral migration rates increase by 1.2 percent over a decade (specifications 1–5).
    - Specification (6) indicates a slightly larger increase equal to 1.5 percent.
  - Differences in temperature-effect estimates are never statistically significant across the six specifications.
- Precipitation, droughts, and floods:
  - Average precipitation:
    - Column (1) coefficient is not statistically significant.
    - Excluding precipitation leaves results unchanged (Column 2).
    - Conclusion: average precipitation is not an important driver of migration in this analysis.
  - Droughts and floods:
    - Coefficients are not statistically significant in the full sample.
    - When using only data from the Global Bilateral Migration Database that ends in 2000 (Column 6), coefficients of deficit and excess of rain become statistically significant and indicate that increases in drought and flood conditions reduce migration rates.
    - Quantified effects (Column 6):
      - A one standard deviation increase in the average measure of drought, which corresponds to 0.3 months in which the SPI is below -1.5 (Table A.1), reduces emigration rates by 7.7 percent.
      - If rain increases by one millimetre during the rainiest days of the year (approximately a 10 percent shock if compared to the sample mean) (Table A.1), emigration rates decrease by 18.7 percent.
    - Context and literature:
      - Empirical evidence on drought effects is mixed (Cattaneo et al. 2019).
      - Prior studies document negative effects of floods on migration in some contexts (e.g., Chen et al. 2017 for Bangladesh; Cattaneo and Massetti 2019 for Nigeria).
      - Floods may induce temporary migration over short distances (Cattaneo et al. 2019).
- Summary conclusion from in-sample results:
  - Temperature is the most robust climatic factor inducing migration in these estimations.

*wpiea2024058-print-pdf - 2000. Migration stocks in the Global Bilateral Migration Database were constructed to be homogeneous*

### conclusion from a meta-analysis that temperature-level changes exert the strongest effects on migration

### Conclusion: Temperature-level changes exert the strongest effects on migration (meta-analysis and study results)

### Key empirical findings
- Meta-analysis cited: Hoffmann et al., 2020 — conclusion that temperature-level changes exert the strongest effects on migration.
- Estimated temperature effects (Table 1, Column (1) baseline PPML): ln(temperature) = 1.168*** (standard error (0.283)). Similar significant positive coefficients in Columns (2) to (6): 1.155***, 1.226***, 1.223***, 1.176***, 1.469*** (standard errors reported in table).
- Precipitation and extreme-event coefficients:
  - ln(precipitation) reported in Table 1 with coefficients including 0.332 and other entries (standard errors shown in table).
  - Drought coefficients include -0.080, -0.109, 0.010, -0.082, -0.256** (standard errors reported).
  - Flood coefficients include -0.082, -0.066, -0.036, -0.083, -0.187** (standard errors reported).
- Robustness checks (Table 2) — OLS and alternate specifications:
  - PPML baseline ln(temp) = 1.168* (standard error (0.699)) in Column (1) of Table 2; other OLS estimates: 0.995, 0.878***, 1.304***, 0.754***, 2.185*** (with standard errors shown).
  - ln(prec) estimates in Table 2 include 0.332, -1.951***, -0.804***, -0.650***, -0.806***, -0.459*** (with standard errors).
  - Drought and Flood retain generally negative coefficients in robustness checks (examples: Drought -0.080, -0.070, -0.095***, -0.252***, -0.094***, -0.336***; Flood -0.082, 0.257***, -0.029*, -0.220***, -0.029**, -0.236***).
- Tests for nonlinearity: Column (5) rejects a nonlinear temperature effect on migration (no evidence of ln(temp)^2 effect), consistent with Cattaneo and Peri 2016.
- First-difference model (Column (6) OLS) still finds migration positively responds to warming while shrinking in response to floods and droughts.
- Potential omitted-variable check for education: adding control for proportion of college graduates (Barro and Lee (2013)) in a restricted sample cannot reject equality of temperature coefficients (t-test = -0.149).

### Historical and projection magnitudes
- Historical total decadal flows baseline: 60 million in 2010.
- Projected total decadal flows across scenarios:
  - 2030: between 73 and 91 million.
  - 2040: between 83 and 102 million.
  - 2050: between 88 and 121 million.
- Uniform warming scenario (+1° to +3°C): total migration increases with warming but projections are never significantly different from the historical average flow at the 95 percent level.
- Climate-only contribution:
  - In most cases, climate change increases outflows of migrants by much less than 5 million per year.
  - When population is held constant, the number of climate migrants is approximately equal to the average number of migrants from 1970 to 2010 (virtually no difference across climate scenarios and decades).
- Comparative literature estimates:
  - Burzyński et al. (2022): additional number of climate migrants from 2010 to 2070 ranges from 37 to 94 million, or 0.6 to 1.6 additional migrants per year, depending on emission scenario (these include sea-level rise, which this study does not).
  - World Bank (Clement et al., 2021): estimates 7 million internal climate migrants per year from 2020 to 2050 (internal migration expected to exceed international migration).
- Regional and bilateral projections (example outcomes for SSP3-4.5 scenario, year 2030):
  - Migration flows from Africa to Europe projected to increase from 3.4 million (2010) to 7 million (2030).
  - Migration flows between Asian countries projected to increase by nearly fifty percent.
  - Flow from Asia to Europe projected to shrink from 5.7 million (2010) to 4.3 million (2030).
  - Flows from Latin America to Europe projected to drop from 3.4 million (2010) to 2.1 million (2030).
  - Flow from Latin America to North America projected to increase by approximately 50 percent.

### Interpretation, caveats, and model limitations
- Exercise is ceteris paribus: assumes only climate and population change; other drivers (socio-economic developments, technology, policy) are held constant.
- Out-of-sample concerns: projecting to 2060 and under RCP 8.5 pushes climate variables beyond historical variance used for estimation; direction of bias is ambiguous:
  - Possible underestimation if large warming triggers disruptive or highly nonlinear migration responses not present historically.
  - Possible positive bias if adaptation reduces migration responses and the model does not capture adaptation.
- Main driver of long-term migration patterns in results: population growth (scenario differences driven primarily by population assumptions).
- Geographic patterns: climate-induced migration increases notably in many African origin countries; intensification of drought and floods can reduce migration in affected countries due to negative estimated coefficients on these extremes.

### Policy implications and recommendations
- Restrictive border policies do not reduce vulnerability to climate change and can limit people’s ability to adapt (cited: Benveniste, Oppenheimer, and Fleurbaey 2020).
- Policies to facilitate temporary migration in the case of extreme events would be beneficial.
- Destination countries concerned about rising migration can pursue policies that foster slower population growth and higher economic development in origin countries (examples: investing in women’s empowerment to boost economic development and slow population growth).
- Migration can act as adaptation: reallocates labor to more productive areas, minimizes health damages and loss of life; receiving countries may gain long-term growth benefits and offset low or negative population dynamics in their own economies.
- Societal impacts in origin and destination countries could be significant and unpredictable; emotional and social costs of migration are potentially large.

*Source: IMF Working Paper — Climate Variability and Worldwide Migration: Empirical Evidence and Projections (conclusion and linked tables/figures).*

### Annex I.  Additional Tables

### Annex I.  Additional Tables

### Table AI.1. Summary Statistics
- Variable Obs. Mean Sd Min Max
- Emigration Rate (*1000) – bilateral 74,437 0.257 3.171 0 215.845
- Emigration Flows – bilateral 74,437 2,408 40,200 0 5,077,120
- Temperature (Degree C) – unilateral 74,437 23.484 5.053 0.082 32.347
- Precipitation (annual mm) - unilateral 74,437 1185 743 28 3377
- Droughts (weighted number of months) - unilateral 74,437 0.225 0.274 0 1.000
- Flood (mm per day) - unilateral 74,437 9.113 6.555 0 27.011

### Origin Countries
- AFG Afghanistan Asia
- AGO Angola Africa
- ALB Albania Europe
- ARE United Arab Emirates Asia
- ARG Argentina Latin America
- BDI Burundi Africa
- BEN Benin Africa
- BFA Burkina Faso Africa
- BGD Bangladesh Asia
- BGR Bulgaria Europe
- BLZ Belize Latin America
- BOL Bolivia Latin America
- BRA Brazil Latin America
- BRN Brunei Asia
- BTN Bhutan Asia
- BWA Botswana Africa
- CAF Central African Republic Africa
- CHL Chile Latin America
- CHN China Asia
- CIV Cote d'Ivoire Africa
- CMR Cameroon Africa
- COG Congo, Rep. Africa
- COL Colombia Latin America
- CRI Costa Rica Latin America
- CUB Cuba Latin America
- DJI Djibouti Africa
- DOM Dominican Republic Latin America
- DZA Algeria Africa
- ECU Ecuador Latin America
- EGY Egypt, Arab Rep. Africa
- ETH Ethiopia (excludes Eritrea) Africa
- GAB Gabon Africa
- GHA Ghana Africa
- GIN Guinea Africa
- GMB Gambia, The Africa
- GNB Guinea-Bissau Africa
- GNQ Equatorial Guinea Africa
- GTM Guatemala Latin America
- GUY Guyana Latin America
- HND Honduras Latin America
- HTI Haiti Latin America
- IDN Indonesia Asia
- IND India Asia
- IRN Iran, Islamic Rep. Asia
- IRQ Iraq Asia
- JOR Jordan Asia
- KEN Kenya Africa
- KHM Cambodia Asia
- KWT Kuwait Asia
- LAO Lao PDR Asia
- LBN Lebanon Asia
- LBR Liberia Africa
- LBY Libya Africa
- LKA Sri Lanka Asia
- LSO Lesotho Africa
- MAR Morocco Africa
- MDG Madagascar Africa
- MEX Mexico Latin America
- MLI Mali Africa
- MMR Myanmar Asia
- MNG Mongolia Asia
- MOZ Mozambique Africa
- MRT Mauritania Africa
- MWI Malawi Africa
- MYS Malaysia Asia
- NER Niger Africa
- NGA Nigeria Africa
- NIC Nicaragua Latin America
- NPL Nepal Asia
- OMN Oman Asia
- PAK Pakistan Asia
- PAN Panama Latin America
- PER Peru Latin America
- PHL Philippines Asia
- PNG Papua New Guinea Oceania
- PRY Paraguay Latin America
- PSE West Bank and Gaza Asia
- RWA Rwanda Africa
- SAU Saudi Arabia Asia
- SEN Senegal Africa
- SLE Sierra Leone Africa
- SLV El Salvador Latin America
- SOM Somalia Africa
- SUR Suriname Latin America
- SWZ Swaziland Africa
- SYR Syrian Arab Republic Asia
- TCD Chad Africa
- TGO Togo Africa
- THA Thailand Asia
- TUN Tunisia Africa
- TZA Tanzania Africa
- UGA Uganda Africa
- URY Uruguay Latin America
- VEN Venezuela Latin America
- VNM Vietnam Asia
- YEM Yemen, Rep. Asia
- ZAF South Africa Africa
- ZAR Congo, Dem. Rep. Africa
- ZMB Zambia Africa
- ZWE Zimbabwe Africa

### Destination Countries
- ABW Aruba Latin America
- AFG Afghanistan Asia
- AGO Angola Africa
- ALB Albania Europe
- ARE United Arab Emirates Asia
- ARG Argentina Latin America
- ATG Antigua and Barbuda Latin America
- AUS Australia Oceania
- AUT Austria Europe
- BDI Burundi Africa
- BEN Benin Africa
- BFA Burkina Faso Africa
- BGD Bangladesh Asia
- BGR Bulgaria Europe
- BHR Bahrain Asia
- BHS Bahamas, The Latin America
- BLZ Belize Latin America
- BOL Bolivia Latin America
- BRA Brazil Latin America
- BRB Barbados Latin America
- BRN Brunei Asia
- BTN Bhutan Asia
- BWA Botswana Africa
- CAF Central African Republic Africa
- CAN Canada North America
- CHE Switzerland Europe
- CHL Chile Latin America
- CHN China Asia
- CIV Cote d'Ivoire Africa
- CMR Cameroon Africa
- COG Congo, Rep. Africa
- COL Colombia Latin America
- COM Comoros Africa
- CPV Cape Verde Africa
- CRI Costa Rica Latin America
- CUB Cuba Latin America
- CYP Cyprus Europe
- DEU Germany Europe
- DJI Djibouti Africa
- DNK Denmark Europe
- DOM Dominican Republic Latin America
- DZA Algeria Africa
- ECU Ecuador Latin America
- EGY Egypt, Arab Rep. Africa
- ESP Spain Europe
- ETH Ethiopia (excludes Eritrea) Africa
- FIN Finland Europe
- FJI Fiji Oceania
- FRA France Europe
- FSM Micronesia, Fed. Sts. Oceania
- GAB Gabon Africa
- GBR United Kingdom Europe
- GHA Ghana Africa
- GIN Guinea Africa
- GLP Guadeloupe Latin America
- GMB Gambia, The Africa
- GNB Guinea-Bissau Africa
- GNQ Equatorial Guinea Africa
- GRC Greece Europe
- GRD Grenada Latin America
- GTM Guatemala Latin America
- GUF French Guiana Latin America
- GUY Guyana Latin America
- HKG Hong Kong, China Asia
- HND Honduras Latin America
- HTI Haiti Latin America
- HUN Hungary Europe
- IDN Indonesia Asia
- IND India Asia
- IRL Ireland Europe
- IRN Iran, Islamic Rep. Asia
- IRQ Iraq Asia
- ISL Iceland Europe
- ISR Israel Asia
- ITA Italy Europe
- JAM Jamaica Latin America
- JOR Jordan Asia
- JPN Japan Asia
- KEN Kenya Africa
- KHM Cambodia Asia
- KIR Kiribati Oceania
- KOR Korea, Rep. Asia
- KWT Kuwait Asia
- LAO Lao PDR Asia
- LBN Lebanon Asia
- LBR Liberia Africa
- LBY Libya Africa
- LCA St. Lucia Latin America
- LKA Sri Lanka Asia
- LSO Lesotho Africa
- LUX Luxembourg Europe
- MAC Macao Asia
- MAR Morocco Africa
- MDG Madagascar Africa
- MDV Maldives Asia
- MEX Mexico Latin America
- MLI Mali Africa
- MLT Malta Europe
- MMR Myanmar Asia
- MNG Mongolia Asia
- MOZ Mozambique Africa
- MRT Mauritania Africa
- MTQ Martinique Latin America
- MUS Mauritius Africa
- MWI Malawi Africa
- MYS Malaysia Asia
- NCL New Caledonia Oceania
- NER Niger Africa
- NGA Nigeria Africa
- NIC Nicaragua Latin America
- NLD Netherlands Europe
- NOR Norway Europe
- NPL Nepal Asia
- NZL New Zealand Oceania
- OMN Oman Asia
- PAK Pakistan Asia
- PAN Panama Latin America
- PER Peru Latin America
- PHL Philippines Asia
- PNG Papua New Guinea Oceania
- POL Poland Europe
- PRI Puerto Rico Latin America
- PRK Korea, Dem. Rep. Asia
- PRT Portugal Europe
- PRY Paraguay Latin America
- PYF French Polynesia Oceania
- QAT Qatar Asia
- REU Reunion Africa
- ROM Romania Europe
- RWA Rwanda Africa
- SAU Saudi Arabia Asia
- SEN Senegal Africa
- SGP Singapore Asia
- SLB Solomon Islands Oceania
- SLE Sierra Leone Africa
- SLV El Salvador Latin America
- SOM Somalia Africa
- STP Sao Tome and Principe Africa
- SUR Suriname Latin America
- SWE Sweden Europe
- SWZ Swaziland Africa
- SYC Seychelles Africa
- SYR Syrian Arab Republic Asia
- TCD Chad Africa
- TGO Togo Africa
- THA Thailand Asia
- TMP East Timor Oceania
- TON Tonga Oceania
- TTO Trinidad and Tobago Latin America
- TUN Tunisia Africa
- TUR Turkey Asia
- TZA Tanzania Africa
- UGA Uganda Africa
- URY Uruguay Latin America
- USA United States North America
- VCT St. Vincent and the Grenadines Latin America
- VEN Venezuela Latin America
- VNM Vietnam Asia
- VUT Vanuatu Oceania
- WSM Samoa Oceania
- YEM Yemen, Rep. Asia
- ZAF South Africa Africa
- ZAR Congo, Dem. Rep. Africa
- ZMB Zambia Africa
- ZWE Zimbabwe Africa

### References (selected from Annex)
- Abel, Guy J. 2018. “Estimates of Global Bilateral Migration Flows by Gender between 1960 and 2015.” International Migration Review 52: 809–52.
- Abel, Guy J., and Joel E. Cohen. 2019. “Bilateral International Migration Flow Estimates for 200 Countries.” Scientific Data 6 (1): 82.
- Alvarez, Jose Luis Cruz, and Esteban Rossi-Hansberg. 2021. “The Economic Geography of Global Warming.” Working Paper 28466. National Bureau of Economic Research.
- Barro, Robert J., and Jong Wha Lee. 2013. “A New Data Set of Educational Attainment in the World, 1950–2010.” Journal of Development Economics 104 (September): 184–98.
- Beine, Michel, Simone Bertoli, and Jesús Fernández-Huertas Moraga. 2016. “A Practitioners’ Guide to Gravity Models of International Migration.” The World Economy 39 (4): 496–512.
- Beine, Michel, and Lionel Jeusette. 2021. “A Meta-Analysis of the Literature on Climate Change and Migration.” Journal of Demographic Economics, 1–52.
- Beine, Michel, and Christopher Parsons. 2015. “Climatic Factors as Determinants of International Migration.” The Scandinavian Journal of Economics 117 (2): 723–67.
- Beine, Michel, and Christopher R. Parsons. 2017. “Climatic Factors as Determinants of International Migration: Redux.” CESifo Economic Studies 63 (4): 386–402.
- Benveniste, Hélène, Michael Oppenheimer, and Marc Fleurbaey. 2020. “Effect of Border Policy on Exposure and Vulnerability to Climate Change.” Proceedings of the National Academy of Sciences 117 (43): 26692–702.
- Benveniste, Hélène, Michael Oppenheimer, and Marc Fleurbaey. 2022. “Climate Change Increases Resource-Constrained International Immobility.” Nature Climate Change 12 (7): 634–41.
- Beyer, Robert M., Jacob Schewe, and Hermann Lotze-Campen. 2022. “Gravity Models Do Not Explain, and Cannot Predict, International Migration Dynamics.” Humanities and Social Sciences Communications 9 (1): 1–10.
- Bohra-Mishra, Pratikshya, Michael Oppenheimer, and Solomon M. Hsiang. 2014. “Nonlinear Permanent Migration Response to Climatic Variations but Minimal Response to Disasters.” Proceedings of the National Academy of Sciences 111 (27): 9780–85.
- Burzynski, Michal, Christoph Deuster, and Frédéric Docquier. 2020. “Geography of Skills and Global Inequality.” Journal of Development Economics 142 (January): 102333.
- Burzyński, Michał, Christoph Deuster, Frédéric Docquier, and Jaime de Melo. 2022. “Climate Change, Inequality, and Human Migration.” Journal of the European Economic Association 20 (3): 1145–97.
- Cai, Ruohong, Shuaizhang Feng, Michael Oppenheimer, and Mariola Pytlikova. 2016. “Climate Variability and International Migration: The Importance of the Agricultural Linkage.” Journal of Environmental Economics and Management 79 (September): 135–51.
- Cattaneo, Cristina, M. Beine, C. Fröhlich, D. Kniveton, I. Martinez-Zarzoso, M. Mastrorillo, K. Millock, E. Piguet, and B. Schraven. 2019. “Human Migration in the Era of Climate Change.” Review of Environmental Economics and Policy.
- Cattaneo, Cristina, and Emanuele Massetti. 2019. “Does Harmful Climate Increase or Decrease Migration? Evidence from Rural Households in Nigeria.” Climate Change Economics 10 (04): 1950013.
- Cattaneo, Cristina, and Giovanni Peri. 2016. “The Migration Response to Increasing Temperatures.” Journal of Development Economics 122: 127–46.
- Chen, Joyce J., Valerie Mueller, Yuanyuan Jia, and Steven Kuo-Hsin Tseng. 2017. “Validating Migration Responses to Flooding Using Satellite and Vital Registration Data.” American Economic Review 107 (5): 441–45.
- CIESIN. 2018. “Gridded Population of the World, Version 4 (GPWv4): Population Density Adjusted to Match 2015 Revision UN WPP Country Totals, Revision 11. NASA Socioeconomic Data and Applications Center (SEDAC).” Center for International Earth Science Information Network, Columbia University.
- Clement, Viviane, Kanta Kumari Rigaud, Alex de Sherbinin, Bryan Jones, Susana Adamo, Jacob Schewe, Nian Sadiq, and Elham Shabahat. 2021. “Groundswell Part 2: Acting on Internal Climate Migration.” Washington, DC: World Bank.
- Dallmann, Ingrid, and Katrin Millock. 2017. “Climate Variability and Inter-State Migration in India.” CESifo Economic Studies 63 (4): 560–94.
- Dao, Thu Hien, Frédéric Docquier, Mathilde Maurel, and Pierre Schaus. 2021. “Global Migration in the Twentieth and Twenty-First Centuries: The Unstoppable Force of Demography.” Review of World Economics 157 (2): 417–49.
- Dell, Melissa, Benjamin F. Jones, and Benjamin A. Olken. 2014. “What Do We Learn from the Weather? The New Climate–Economy Literature.” Journal of Economic Literature 52 (3): 740–98.
- Desmet, Klaus, Robert E. Kopp, Scott A. Kulp, Dávid Krisztián Nagy, Michael Oppenheimer, Esteban Rossi-Hansberg, and Benjamin H. Strauss. 2021. “Evaluating the Economic Cost of Coastal Flooding.” American Economic Journal: Macroeconomics 13 (2): 444–86.
- Desmet, Klaus, Dávid Krisztián Nagy, and Esteban Rossi-Hansberg. 2018. “The Geography of Development.” Journal of Political Economy 126 (3): 903–83.
- Desmet, Klaus, and Esteban Rossi-Hansberg. 2015. “On the Spatial Economic Impact of Global Warming.” Journal of Urban Economics 88 (July): 16–37.
- Diaz, Delavane B. 2016. “Estimating Global Damages from Sea Level Rise with the Coastal Impact and Adaptation Model (CIAM).” Climatic Change 137 (1): 143–56.
- Fagan, Brian. 2008. The Great Warming: Climate Change and the Rise and Fall of Civilizations.
- FAO, and IIASA. 2022. “Global Agro Ecological Zones Version 4 (GAEZ V4).”
- Felbermayr, Gabriel, and Jasmin Gröschl. 2014. “Naturally Negative: The Growth Effects of Natural Disasters.” Journal of Development Economics 111 (C): 92–106.
- Gröschl, Jasmin, and Thomas Steinwachs. 2017. “Do Natural Hazards Cause International Migration?*.” CESifo Economic Studies 63 (4): 445–80.
- Hoffmann, Roman, Anna Dimitrova, Raya Muttarak, Jesus Crespo Cuaresma, and Jonas Peisker. 2020. “A Meta-Analysis of Country-Level Studies on Environmental Change and Migration.” Nature Climate Change 10 (10): 904–12.
- Hsiang, Solomon. 2016. “Climate Econometrics.” Annual Review of Resource Economics 8 (1): 43–75.
- Ionesco, Dina, Daria Mokhnacheva, and François Gemenne. 2016. The Atlas of Environmental Migration. 1° edizione.
- IPCC. 2022. “Climate Change 2022: Impacts, Adaptation, and Vulnerability. Contribution of Working Group II to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change.”
- Jessoe, Katrina, Dale T. Manning, and J. Edward Taylor. 2017. “Climate Change and Labour Allocation in Rural Mexico: Evidence from Annual Fluctuations in Weather.” The Economic Journal, June.
- Kahn, Matthew E. 2010. Climatopolis: How Our Cities Will Thrive in a Hotter Future.
- Lachniet, Matthew S., Yemane Asmerom, Victor Polyak, and Juan Pablo Bernal. 2017. “Two Millennia of Mesoamerican Monsoon Variability Driven by Pacific and Atlantic Synergistic Forcing.” Quaternary Science Reviews 155 (January): 100–113.
- Marchiori, Luca, Jean-François Maystadt, and Ingmar Schumacher. 2012. “The Impact of Weather Anomalies on Migration in Sub-Saharan Africa.” Journal of Environmental Economics and Management 63 (3): 355–74.
- McGuirk, Eoin, and Marshall Burke. 2020. “The Economic Origins of Conflict in Africa.” Journal of Political Economy, September.
- McKee, Thomas B., Nolan J. Doesken, and John Kleist. 1993. “The Relationship of Drought Frequency and Duration to Time Scales.” 8th Conference on Applied Climatology, Anaheim, 179–84.
- McMichael, Celia, Jon Barnett, and Anthony J. McMichael. 2012. “An Ill Wind? Climate Change, Migration, and Health.” Environmental Health Perspectives 120 (5): 646–54.
- Missirian, Anouch, and Wolfram Schlenker. 2017. “Asylum Applications Respond to Temperature Fluctuations.” Science 358 (6370): 1610–14.
- Mueller, V., C. Gray, and K. Kosec. 2014. “Heat Stress Increases Long-Term Human Migration in Rural Pakistan.” Nature Climate Change 4 (3): 182–85.
- Myers, Norman. 1993. “Environmental Refugees in a Globally Warmed World.” BioScience 43 (11): 752–61.
- Myers, Norman. 2002. “Environmental Refugees: A Growing Phenomenon of the 21st Century.” Philosophical Transactions of the Royal Society B: Biological Sciences 357 (1420): 609–13.
- O’Neill, Brian C., Elmar Kriegler, Kristie L. Ebi, Eric Kemp-Benedict, Keywan Riahi, Dale S. Rothman, Bas J. van Ruijven, et al. 2017. “The Roads Ahead: Narratives for Shared Socioeconomic Pathways Describing World Futures in the 21st Century.” Global Environmental Change 42: 169–80.
- Ozden, C., C. R. Parsons, M. Schiff, and T. L. Walmsley. 2011. “Where on Earth Is Everybody? The Evolution of Global Bilateral Migration 1960-2000.” The World Bank Economic Review 25 (1): 12–56.
- Peri, Giovanni, and Akira Sasahara. 2019. “The Impact of Global Warming on Rural-Urban Migrations: Evidence from Global Big Data.” Working Paper 25728. National Bureau of Economic Research.
- Rigaud, K. K., A. de Sherbinin, B. Jones, J. Bergmann, V. Clement, K. Ober, J. Schewe, et al. 2018. Groundswell: Preparing for Internal Climate Migration. World Bank, Washington, DC.
- Rodell, M., P. R. Houser, U. Jambor, J. Gottschalck, K. Mitchell, C.-J. Meng, K. Arsenault, et al. 2004. “The Global Land Data Assimilation System.” Bulletin of the American Meteorological Society 85 (3): 381–94.
- Santos Silva, J. M. C. Santos, and Silvana Tenreyro. 2006. “The Log of Gravity.” Review of Economics and Statistics 88 (4): 641–58.
- Sherbinin, A. de de, M. Castro, F. Gemenne, M. M. Cernea, S. Adamo, P. M. Fearnside, G. Krieger, et al. 2011. “Preparing for Resettlement Associated with Climate Change.” Science 334 (6055): 456–57.
- Stern, Nicholas. 2007. The Economics of Climate Change: The Stern Review. Cambridge: Cambridge University Press.
- Thrasher, B., E. P. Maurer, C. McKellar, and P. B. Duffy. 2012. “Technical Note: Bias Correcting Climate Model Simulated Daily Temperature Extremes with Quantile Mapping.” Hydrology and Earth System Sciences 16 (9): 3309–14.
- United Nations Population Division. 2015a. “Trends in International Migrant Stock: The 2015 Revision.” United Nations, Department of Economic and Social Affairs, New York, NY.
- United Nations Population Division. 2015b. “World Population Prospects: The 2015 Revision.” United Nations, Department of Economic and Social Affairs, New York, NY.
- Vicente-Serrano, Sergio M., Santiago Beguería, and Juan I. López-Moreno. 2010. “A Multiscalar Drought Index Sensitive to Global Warming: The Standardized Precipitation Evapotranspiration Index.” Journal of Climate 23 (7): 1696–1718.
- Weiss, Harvey, and Raymond S. Bradley. 2001. “What Drives Societal Collapse?” Science 291 (5504): 609–10.

*Climate Variability and Worldwide Migration: Empirical Evidence and Projections. Working Paper No. WP/2024/058*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024058-print-pdf.pdf_