## Executive Summary — The Heterogeneous Effects of Uncertainty on Trade ## Source details **Canonical URL:** [Executive Summary — The Heterogeneous Effects of Uncertainty on Trade](https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024139-print-pdf.pdf) ## Other formats - [Markdown version](/-/media/files/publications/wp/2024/english/wpiea2024139-print-pdf.pdf.md) - [Structured JSON version](/-/media/files/publications/wp/2024/english/wpiea2024139-print-pdf.pdf.json) --- ### Key findings - Sample and method - Sample: 143 countries over the 1980-2021 period. - Empirical framework: augmented gravity model estimated using Poisson Pseudo-Maximum Likelihood (PPML). - Uncertainty measure: text-based index from Ahir, Bloom, and Furceri (2018, 2022) constructed from the frequency of the word “uncertainty” in Economist Intelligence Unit (EIU) country reports (indices normalized by total words and rescaled by multiplying by 1,000). - Main quantitative result - A one standard deviation increase in global uncertainty is associated with a decline in bilateral trade by 4.5 percent. - Baseline coefficients on aggregated and disaggregated uncertainty indices are negative and strongly significant at the 1 percent level. - Sectoral and directionality effects - Mineral fuel trade and manufacturing products trade are the most impacted by uncertainty. - The negative impact is observed for uncertainty originating on both sides of the border, with a higher impact from uncertainty in the importing country (demand effect > supply effect). - Heterogeneity across partners and regions - Uncertainty reduces bilateral trade between Emerging Markets and Developing Economies (EMDEs) and Advanced Economies (AEs). - Intraregional trade declines in Africa and Europe as uncertainty surges. - Non-resources-rich countries are more at risk from uncertainty shocks. - Integration, global value chains, and geopolitics - Horizontal trade integration mitigates the negative impact of uncertainty on trade. - Vertical integration (higher participation in Global Value Chains — GVCs) amplifies the negative impact of uncertainty on trade. - Geopolitical tensions and higher geopolitical risk amplify the deterrent effect of uncertainty on trade. - Stronger diplomatic links and lower geopolitical risk help contain the negative impact. ### Methodology and data highlights - Gravity model specification - Bilateral trade flows Xij modeled as function of exporter and importer GDP, bilateral trade costs (distance, common language, colonial ties, contiguity), trade agreements, and uncertainty terms for exporter and importer. - Joint bilateral uncertainty defined as the sum of uncertainty from exporting and importing countries. - Estimation choices and rationale - PPML estimator used to handle zero bilateral trade observations and heteroskedasticity; standardized uncertainty coefficients are used to interpret effects as percent changes in trade per one standard deviation (SD) change in uncertainty. - Data sources (selected) - Trade, exports, imports: CEPII-BACI database, IMF, UNCTAD. - Geographic variables and diplomatic disagreement: CEPII. - Uncertainty: Ahir, Bloom and Furceri (2018, 2022). - Geopolitical Risk: Caldara and Iacoviello (2022). - GVC participation: EORA-MRIO databases. - Trade agreements: Mario Larch’s RTA Database. - Additional contextual metrics cited - Red Sea Houthi Attacks (started on November 19, 2023) — 40 percent of Asia-Europe trade channels through the region. - Hormuz Strait channeled 21 percent of global petroleum liquid consumption. - Panama Canal drought — about 5 percent of global maritime trade volumes ship through the Panama Canal. ### Main empirical results — magnitudes and directionality - Joint and country-specific uncertainty - A one SD hike in joint uncertainty is associated with a decline in bilateral trade by 3.85 percent (column [1]). - A one SD increase in uncertainty (exporter or importer) is equivalent to a level of uncertainty of 0.29 (example: Madagascar in 2017). - A one SD increase in joint uncertainty is equivalent to an increase in joint uncertainty by 0.52 (examples: Russia–Bangladesh 2004; Russia–Greece 2021; Guinea–Côte d’Ivoire 2008). - Asymmetry: importing vs exporting country uncertainty - A one SD increase in global uncertainty in the importing country is associated with a decline in bilateral trade by 2.6 percent (Table 2, column [2]). - A one SD increase in global uncertainty in the exporting country is associated with a decline in bilateral trade by 2.4 percent (Table 2, column [2]). - Lagged uncertainty and endogeneity checks - A one SD increase in lagged joint uncertainty is associated with a decline in bilateral trade by 4.5 percent (Table 2, column [3]). - A one SD increase in lagged uncertainty in the importer country → decline of 3.1 percent (0.57 percentage points higher than baseline). - A one SD increase in lagged uncertainty in the exporter country → decline of 2.8 percent (Table 2, column [4]). - Table 2 summary: Observations 425,993 (columns [1] and [2]); 419,455 (columns [3] and [4]). R2: 0.9735 and 0.9741 respectively. Fixed effects: Exporter, Importer, Pair, Time. - Supply vs demand channels (Table 3) - Lagged exporter uncertainty (supply effect): coefficient -0.162*** (standard error 0.0450); beta [-0.0229]. - Lagged importer uncertainty (demand effect): coefficient -0.219*** (standard error 0.0317); beta [-0.0310]. - Observations: 419,455 (supply), 419,454 (demand). R2: 0.9807 in both columns. - Interpretation: demand effects (importer uncertainty) have larger negative impact than supply effects. ### Sectoral heterogeneity - Primary products (lagged joint uncertainty) - Joint Uncertainty t−1 coefficient: -0.122*** (0.0337); beta [-0.0266]. - One SD increase of lagged joint uncertainty → primary products trade decline by 2.7 percent. - Exporter uncertainty t−1 → decline by 1.3 percent; importer uncertainty t−1 → decline by 2.2 percent. - Mineral fuel (lagged joint uncertainty) - Joint Uncertainty t−1 coefficient: -0.190*** (0.0535); beta [-0.0414]. - One SD increase of lagged joint uncertainty → mineral fuel trade decline by 4.1 percent. - Exporter uncertainty t−1 → decline by 2.9 percent; importer uncertainty t−1 → decline by 2.5 percent. - Interpretation: supply effects relatively more important for mineral fuel. - Manufacturing products (lagged joint uncertainty) - Joint Uncertainty t−1 coefficient: -0.233*** (0.0343); beta [-0.0507]. - One SD increase of lagged joint uncertainty → manufacturing trade decline by 5.1 percent. - Exporter uncertainty t−1 → decline by 3.2 percent; importer uncertainty t−1 → decline by 3.4 percent. ### Natural resource endowment heterogeneity - Sample split by Sawadogo (2020) dummy: resource-rich vs non-resource-rich. - Lagged uncertainty in natural resource-rich countries generally does not significantly impact their bilateral trade. - Non-natural resource-rich countries more negatively affected: - One SD hike in lagged joint uncertainty in non-natural resource-rich countries reduces bilateral trade by 4.4 percent (Table 5, column [7]). - Exporter non-resource-rich & importer resource-rich: one SD increase → bilateral trade decreases by 5.9 percent (Table 5, column [6]). - Both importer and exporter non-resource-rich: - Exporter uncertainty one SD → decreases bilateral trade by 2.7 percent. - Importer uncertainty one SD → decreases bilateral trade by 3.0 percent (Table 5, column [8]). - Observations by subsample: Exporter->Resource-rich / Importer->Resource-rich: 29,186; Exporter->Non-Resource-rich / Importer->Non-Resource-rich: 212,369. - R2 range: 0.9645 to 0.9790. ### Development-stage heterogeneity (EMDEs vs AEs) - Lagged joint uncertainty effects: - Intra-EMDEs decreases trade by 3.3 percent (Table 6, column [1]). - Intra-AEs decreases trade by 2.7 percent (Table 6, column [7]). - Joint uncertainty exporting EMDE → importing AE decreases trade by 7.2 percent (Table 6, column [3]). - Joint uncertainty exporting AE → importing EMDE decreases trade by 2.4 percent (Table 6, column [5]). - Directional examples: - Exporter AE → Importer EMDE: - One SD increase of lagged exporter AE uncertainty decreases bilateral trade with EMDE by 1.6 percent. - One SD increase of lagged importer EMDE uncertainty decreases bilateral trade with AE by 1.5 percent (Table 6, column [4]). - Exporter EMDE → Importer AE: - One SD increase of lagged exporter EMDE uncertainty decreases bilateral trade with AE by 4.3 percent. - Selected coefficients (Table 6; beta coefficients in brackets): - Joint Uncertainty t−1: Column [1] -0.121*** (0.0302) [-0.0265]; Column [3] -0.328*** (0.0477) [-0.0716]. - Uncertainty from exporter t−1: Column [3] -0.303*** (0.0581) [-0.0430]. - Uncertainty from importer t−1: Column [3] -0.367*** (0.0775) [-0.0519]. - Observations (selected): 32,718; 113,448; 79,887; 193,402. R2 examples: 0.967; 0.900; 0.842; 0.798. ### Intraregional vs extra-continental trade - Intraregional results - Intra-African and intra-European trade are most negatively affected. - A one SD increase of uncertainty in exporting countries: - Decreases intra-African bilateral trade by 4.2 percent. - Decreases intra-European bilateral trade by 2.1 percent. - For intra-European trade, a one SD increase of importer uncertainty decreases trade by 1.2 percent (Table 7, columns [2] and [8]). - Extra-continental results - About half of the coefficients in front of lagged joint uncertainty are negative and significant at the 1 percent level (Table A4). - For Africa: one SD hike in lagged joint uncertainty decreases bilateral trade between exporting African countries and the rest of the world (ROW); increase in lagged joint uncertainty for importing African country and ROW does not significantly impact trade flows. - Table 7 selected intracontinental coefficients (beta coefficients in brackets): - Joint Uncertainty t−1: Africa -0.114 (0.0750) [-0.0248]; Europe -0.115*** (0.0325) [-0.0251]. - Uncertainty from exporter t−1: Africa -0.295*** (0.0985) [-0.0418]; Europe -0.145*** (0.0466) [-0.0206]. - Observations examples: Africa 28,242; Asia 33,521; Europe 32,243. R2 examples: 0.750; 0.880; 0.952. ### Conditional factors — trade integration, GVCs, and geopolitics - Horizontal integration (trade intensity) - A one SD increase in lagged joint uncertainty: - Decreases bilateral trade by approximately 7.3 percent for country pairs at the 25th percentile of trade intensity. - Decreases bilateral trade by 0.37 percent for pairs at the 70th percentile of trade intensity. - Impact becomes positive at the 75th percentile of trade intensity. - Exporter uncertainty: one SD → approximately 6.8 percent decrease at 25th percentile; 1.1 percent decrease at 75th percentile; impact becomes positive at 85th percentile. - Importer uncertainty: one SD → 2.8 percent decrease at 25th percentile; 1.5 percent increase at 75th percentile. - Interpretation: deeper bilateral trade intensity mitigates negative uncertainty effects. - Vertical integration (GVC participation) - A one SD increase in lagged joint uncertainty: - Decreases bilateral trade by 2.6 percent at 25th percentile of joint GVC participation. - Decreases bilateral trade by 5.6 percent at 75th percentile of joint GVC participation. - Exporter GVC participation: one SD exporter uncertainty → ~2 percent decrease at 25th percentile; ~6 percent decrease at 75th percentile. - Importer GVC participation: one SD importer uncertainty → ~0.1 percent decrease at 25th percentile; ~2.5 percent decrease at 75th percentile. - Interpretation: higher GVC involvement, especially for exporters, amplifies negative impact. - Geopolitical risk - Joint geopolitical risk ranges 0 to 6.63; country-level ranges 0 to 4.35. - A one SD hike in lagged joint uncertainty: - Decreases bilateral trade by 1.9 percent for partners with joint geopolitical risk = 1. - Decreases bilateral trade by 9.6 percent for partners with joint geopolitical risk = 3. - Exporter geopolitical risk examples: - Geopolitical risk = 0: one SD exporter uncertainty → ~0.3 percent increase in bilateral trade. - Geopolitical risk = 1: one SD exporter uncertainty → 6.8 percent decrease. - Geopolitical risk = 2: one SD exporter uncertainty → 14 percent decrease. - Interpretation: higher geopolitical risk deepens uncertainty’s negative effect. - Diplomatic disagreement (bilateral geopolitical distance) - A one SD increase in lagged joint uncertainty: - Leads to ~0.067 percent decrease in bilateral trade at 25th percentile of diplomatic disagreement (friendly pairs). - Leads to 6.4 percent decrease at 75th percentile (high diplomatic rivals). - Importer uncertainty: one SD hike → 1 percent increase in bilateral trade between countries at 25th percentile of diplomatic disagreement; 2.8 percent decrease between nonfriendly countries. - Interpretation: diplomatic rivalry amplifies how uncertainty reduces bilateral trade. ### Robustness and alternative uncertainty measures - Mean, Max, Min, and difference measures (Table 8) - Mean_Uncertainty_t-1 coefficient: -0.353*** (0.0619) [beta: -0.0386]; Observations: 426,578; R2: 0.9735. - Maximum Uncertainty_t-1 coefficient: -0.180*** (0.0418) [beta: -0.0285]; Observations: 426,578; R2: 0.9734. - Minimum Uncertainty_t-1 coefficient: -0.403*** (0.0641) [beta: -0.0343]; Observations: 426,578; R2: 0.9735. - Difference and Absolute Difference measures are not statistically significant. - Additional controls (volatility of credit, terms of trade, industrial value added, bilateral exchange rate) - A 1 SD hike in lagged joint uncertainty reduces trade by 3.1 percent (Table A9, column [1]). - Lagged importer uncertainty one SD → decline of 2.24 percent (Table A9, column [2]). - Lagged exporter uncertainty one SD → decline of 1.79 percent (Table A9, column [2]). - Robustness checks indicate demand effects are more important than supply effects. ### Other types and sources of uncertainty - Trade-related and pandemic-related uncertainty - World Trade Uncertainty Index shows highest hike during US-China trade tensions. - World Pandemic Uncertainty Index records highest episodes during SARS (2002–2003) and COVID-19 (2020–23). - Trade uncertainty: one SD increase in lagged joint trade uncertainty → reduction in bilateral trade of 1.6 percent (Table 9, column [1]). - Pandemic uncertainty: one SD increase in lagged joint pandemic uncertainty → reduction in bilateral trade of 4.6 percent (Table 10, column [1]). - Trade uncertainty by source: - Exporter trade uncertainty one SD → reduces bilateral trade by 1.3 percent. - Importer trade uncertainty one SD → reduces bilateral trade by 0.7 percent. - Interpretation: trade uncertainty shows a higher supply effect. - Pandemic uncertainty by source: - Importer pandemic uncertainty one SD → reduces bilateral trade by 2.6 percent. - Exporter pandemic uncertainty one SD → reduces bilateral trade by 2.5 percent. - Interpretation: pandemic-related demand effects slightly more important than supply effects. ### GMM methodology and additional endogeneity checks - System-GMM and IV/GMM approaches employed to address endogeneity; instruments include lagged uncertainty and second-order lagged log trade. - Main GMM estimates summary (Table 11) - IV/GMM Joint Uncertainty: -0.104*** (0.0230) [ -0.0227 ] (column [1]). - System GMM Joint Uncertainty: -0.211*** (0.0753) [ -0.0477 ] (column [3]). - IV/GMM exporter uncertainty: -0.145*** (0.0330) [ -0.0205 ]. - IV/GMM importer uncertainty: -0.0644** (0.0324) [ -0.00912 ]. - Lagged log trade coefficients: 0.772*** (0.00515) in IV/GMM; 0.577*** (0.0144) in system GMM. - Observations: 354,443 (IV/GMM columns [1]/[2]); 60,664 (System GMM columns [3]/[4]). - Magnitude interpretation from GMM - A one SD increase in joint uncertainty reduces bilateral trade by 2.3 percent (Table 11, column [1]). - A one SD increase in exporter uncertainty → decline of 2.1 percent. - A one SD increase in importer uncertainty → decline of 1.0 percent. - System GMM results yield higher coefficients, consistent with baseline. ### Policy implications and recommendations - Twofold policy message - Trade is sensitive to high uncertainty shocks regardless of the origin of the uncertainty; such shocks can weaken conventional economic policies and sabotage past development gains, especially in EMDEs that rely on trade for basic needs. - Policymakers should distinguish between horizontal and vertical integration when designing safety nets: horizontal integration attenuates uncertainty’s negative effect on trade, while vertical (GVC) integration exacerbates it. - Practical policy considerations - Design trade policies and protocols to prevent economic losses from potential uncertainty shocks and geopolitical fragmentation. - Anticipate local uncertainty shocks from trading partners and assess geopolitical relations to determine appropriate responses. - Anticipate geopolitical risks in trading partners to design targeted responses and resilience measures (e.g., diversification, contingency protocols for critical shipping routes and energy supplies). - Promote openness and trade integration while complementing it with measures to prevent economic losses from uncertainty shocks; highly integrated partners can become unavoidable trading partners that help counter negative impacts. *Source: IMF Working Paper — Executive Summary (wpiea2024139-print-pdf).* ### Executive Summary ...................................................................................................... ### Executive Summary ### Major themes and structure - Executive Summary (page 3) - I. Introduction (page 4) - II. Literature review (page 5) - III. Methodology and Data (page 7) - A. Conceptual framework of the gravity model (page 7) - B. Data (page 8) - IV. Results (page 10) - A. Baseline results (page 10) - B. Extensions: deep dive (page 13) - 1. Sectoral trade and resource endowment (page 13) - 2. Heterogenous impact (page 17) - 3. Conditional factors (page 20) - C. Robustness (page 24) - 1. Alternative measures of global uncertainty and additional controls (page 24) - 2. Other types of uncertainty (page 25) - 3. GMM methodology (page 27) - V. Conclusion (page 28) - References (page 39) - Title on page header: IMF WORKING PAPERS The Heterogeneous Effects of Uncertainty on Trade (page 3) *Source: wpiea2024139-print-pdf - Executive Summary* ### Executive Summary ### Executive Summary — The Heterogeneous Effects of Uncertainty on Trade ### Key findings - Sample and method - Sample: 143 countries over the 1980-2021 period. - Empirical framework: augmented gravity model estimated using Poisson Pseudo-Maximum Likelihood (PPML). - Uncertainty measure: text-based index from Ahir, Bloom, and Furceri (2018, 2022) constructed from the frequency of the word “uncertainty” in Economist Intelligence Unit (EIU) country reports (indices normalized by total words and rescaled by multiplying by 1,000). - Main quantitative result - A one standard deviation increase in global uncertainty is associated with a decline in bilateral trade by 4.5 percent. - Baseline coefficients on aggregated and disaggregated uncertainty indices are negative and strongly significant at the 1 percent level. - Sectoral and directionality effects - Mineral fuel trade and manufacturing products trade are the most impacted by uncertainty. - The negative impact is observed for uncertainty originating on both sides of the border, with a higher impact from uncertainty in the importing country (demand effect > supply effect). - Heterogeneity across partners and regions - Uncertainty reduces bilateral trade between Emerging Markets and Developing Economies (EMDEs) and Advanced Economies (AEs). - Intraregional trade declines in Africa and Europe as uncertainty surges. - Non-resources-rich countries are more at risk from uncertainty shocks. - Integration, global value chains, and geopolitics - Horizontal trade integration (deeper bilateral trade intensity) mitigates the negative impact of uncertainty on trade. - Vertical integration (higher participation in Global Value Chains — GVCs) amplifies the negative impact of uncertainty on trade. - Geopolitical tensions and higher geopolitical risk amplify the deterrent effect of uncertainty on trade. - Stronger diplomatic links and lower geopolitical risk help contain the negative impact. ### Methodology and data highlights - Gravity model specification - Bilateral trade flows Xij modeled as function of exporter and importer GDP, bilateral trade costs (distance, common language, colonial ties, contiguity), trade agreements, and uncertainty terms for exporter and importer. - Joint bilateral uncertainty defined as the sum of uncertainty from exporting and importing countries. - Estimation choices and rationale - PPML estimator used to handle zero bilateral trade observations and heteroskedasticity; standardized uncertainty coefficients are used to interpret effects as percent changes in trade per one standard deviation (SD) change in uncertainty. - Data sources (selected) - Trade, exports, imports: CEPII-BACI database, IMF, UNCTAD. - Geographic variables and diplomatic disagreement: CEPII. - Uncertainty: Ahir, Bloom and Furceri (2018, 2022). - Geopolitical Risk: Caldara and Iacoviello (2022). - GVC participation: EORA-MRIO databases. - Trade agreements: Mario Larch’s RTA Database. - Additional contextual metrics cited - Recent events that raise uncertainty and can affect trade: Red Sea Houthi Attacks (started on November 19, 2023) — 40 percent of Asia-Europe trade channels through the region; Hormuz Strait channeled 21 percent of global petroleum liquid consumption; Panama Canal drought — about 5 percent of global maritime trade volumes ship through the Panama Canal. ### Robustness and additional results - Results robust to alternative measures of uncertainty, inclusion of additional control variables, and estimation strategy variations. - Nonlinearities and interaction effects explored: - Trade intensity and GVC participation used as interactive variables to capture moderating or amplifying roles. - Geopolitical risk and diplomatic disagreement interact with uncertainty to deepen adverse trade effects. ### Policy implications and recommendations - Twofold policy message - Trade is sensitive to high uncertainty shocks regardless of the origin of the uncertainty; such shocks can weaken conventional economic policies and sabotage past development gains, especially in EMDEs that rely on trade for basic needs. - Policymakers should distinguish between horizontal and vertical integration when designing safety nets: horizontal integration attenuates uncertainty’s negative effect on trade, while vertical (GVC) integration exacerbates it. - Practical policy considerations - Design trade policies and protocols to prevent economic losses from potential uncertainty shocks and geopolitical fragmentation. - Anticipate local uncertainty shocks from trading partners and assess geopolitical relations to determine appropriate responses. - Anticipate geopolitical risks in trading partners to design targeted responses and resilience measures (e.g., diversification, contingency protocols for critical shipping routes and energy supplies). *Source: IMF Working Paper — Executive Summary (wpiea2024139-print-pdf).* ### 3.85 percent (column [1]). This finding is slightly similar to previous findings by Jakubik and Ruta (2023), ### wpiea2024139-print-pdf - 3.85 percent (column [1]). This finding is slightly similar to previous findings by Jakubik and Ruta (2023), ### Impact of joint and country-specific uncertainty on bilateral trade - A one SD hike in joint uncertainty is associated with a decline in bilateral trade by 3.85 percent (column [1]). - Comparable prior finding: a one SD hike in global trade policy uncertainty leads to a 3.1 percent decline in bilateral trade between geopolitical rival countries (Jakubik and Ruta (2023)). - A one SD increase in joint uncertainty is equivalent to an increase in joint uncertainty by 0.52, which is equivalent to the level of joint uncertainty between Russia and Bangladesh in 2004; Russia and Greece in 2021, or Guinea and Côte d’Ivoire in 2008. - A one SD increase in uncertainty (exporter or importer) is equivalent to a level of uncertainty of 0.29, which is equivalent to the level of uncertainty in Madagascar in 2017. ### Asymmetry: importing vs exporting country uncertainty - Uncertainty originating from the importing country is more detrimental than uncertainty from the exporting country. - A one SD increase in global uncertainty in the importing country is associated with a decline in bilateral trade by 2.6 percent (Table 2, column [2]). - A one SD increase in global uncertainty in the exporting country is associated with a decline in bilateral trade by 2.4 percent (Table 2, column [2]). - Explanation: importing-country uncertainty affects demand (income/capacity to import) and product-type irreducibility; exporting-country uncertainty can restrict exports through reduced production capacity (Pangestu & Trotsenburg, 2022). ### Lagged uncertainty and endogeneity checks - To address potential reverse causality, the study uses lagged uncertainty. - Lagging uncertainty confirms and slightly deepens the negative impacts: - A one SD increase in lagged joint uncertainty is associated with a decline in bilateral trade by 4.5 percent (Table 2, column [3]). - A one SD increase in lagged uncertainty in the importer country is associated with a decline in bilateral trade of 3.1 percent, which is 0.57 percent higher than the baseline result. - A one SD increase of lagged uncertainty in the exporter country is associated with a decline in bilateral trade by 2.8 percent (Table 2, column [4]). - All coefficients associated with different uncertainty measures (lagged) are negative and strongly significant at the 1 percent level (PPML approach). - Table 2 summary statistics: - Observations: 425,993 (columns [1] and [2]); 419,455 (columns [3] and [4]) - R2: 0.9735 (columns [1] and [2]); 0.9741 (columns [3] and [4]) - Fixed effects: Exporter, Importer, Pair, Time - Yes in all specifications. - Note: Beta coefficients shown between brackets. Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. ### Supply vs demand effects (Table 3) - Estimation strategy controls for exporter-year and importer-year fixed effects to separate supply and demand channels. - Lagged uncertainty from exporter (supply effect): - Coefficient: -0.162*** (standard error 0.0450); beta coefficient [-0.0229]. - Lagged uncertainty from importer (demand effect): - Coefficient: -0.219*** (standard error 0.0317); beta coefficient [-0.0310]. - Observations: 419,455 (supply), 419,454 (demand). - R2: 0.9807 in both columns. - Fixed effects: Exporter, Importer, Pair, Time present; Exporter-Year fixed effects included when assessing importer effects and vice versa. - Interpretation: demand effects (uncertainty in importer) have larger negative impact than supply effects (uncertainty in exporter). ### Sectoral heterogeneity (Table 4; Figure 3) - Aggregate finding: impact of uncertainty varies by sector; manufacturing and mineral fuel are more sensitive. - Primary products (lagged joint uncertainty): - Joint Uncertainty t−1 coefficient: -0.122*** (0.0337); beta [-0.0266]. - Uncertainty from exporter t−1: -0.0922* (0.0501); beta [-0.0131]. - Uncertainty from importer t−1: -0.157*** (0.0438); beta [-0.0223]. - One SD increase of lagged joint uncertainty → primary products trade decline by 2.7 percent (Table 4, column [1]). - One SD increase of lagged uncertainty from exporter → primary products decline by 1.3 percent; from importer → 2.2 percent (Table 4, columns [2]). - Mineral fuel (lagged joint uncertainty): - Joint Uncertainty t−1 coefficient: -0.190*** (0.0535); beta [-0.0414]. - Uncertainty from exporter t−1: -0.203** (0.0831); beta [-0.0287]. - Uncertainty from importer t−1: -0.176*** (0.0622); beta [-0.0250]. - One SD increase of lagged joint uncertainty → mineral fuel trade decline by 4.1 percent (Table 4, column [5]). - One SD increase of lagged uncertainty in exporter → mineral fuel decline by 2.9 percent; in importer → 2.5 percent (Table 4, column [6]). - Interpretation: supply effects are relatively more important for mineral fuel (uncertainty in oil-exporting countries decreases oil supply and increases prices). - Manufacturing products (lagged joint uncertainty): - Joint Uncertainty t−1 coefficient: -0.233*** (0.0343); beta [-0.0507]. - Uncertainty from exporter t−1: -0.226*** (0.0565); beta [-0.0320]. - Uncertainty from importer t−1: -0.239*** (0.0464); beta [-0.0338]. - One SD increase of lagged joint uncertainty → manufacturing trade decline by 5.1 percent (Table 4, columns [7]). - One SD increase of lagged uncertainty in exporter → manufacturing decline by 3.2 percent; in importer → 3.4 percent (Table 4, columns [8]). - Interpretation: manufacturing trade declines most strongly; discretionary/luxury nature and reducible consumption contribute. ### Natural resource endowment heterogeneity (Table 5) - Sample split between natural resource-rich and non-natural resource-rich countries (Sawadogo (2020) dummy). - Lagged uncertainty in natural resource-rich countries generally does not significantly impact their bilateral trade. - Bilateral trade involving non-natural resource-rich countries is more negatively affected: - One SD hike in lagged joint uncertainty in non-natural resource-rich countries reduces their bilateral trade by 4.4 percent (Table 5, column [7]). - When exporting country is non-resource-rich and importing country resource-rich: one SD increase in lagged uncertainty decreases bilateral trade by 5.9 percent (Table 5, column [6]). - When both importer and exporter are non-natural resource-rich: - One SD increase of lagged uncertainty in exporting country decreases bilateral trade by 2.7 percent. - One SD increase of lagged uncertainty in importing country decreases bilateral trade by 3 percent (Table 5, column [8]). - Observations by subsample: - Exporter->Resource-rich / Importer->Resource-rich: 29,186 observations. - Exporter->Non-Resource-rich / Importer->Non-Resource-rich: 212,369 observations. - R2 range: 0.9645 to 0.9790 across columns. - Interpretation: uncertainty emanating from non-resource-rich countries has larger negative trade effects globally. ### Development-stage heterogeneity (EMDEs vs AEs; Table 6 summary) - Uncertainty in EMDEs has the highest impact on trade. - Lagged joint uncertainty: - Intra-EMDEs decreases trade by 3.3 percent (Table 6, column [1]). - Intra-AEs decreases trade by 2.7 percent (Table 6, column [7]). - Joint uncertainty between exporting EMDEs and importing AEs decreases trade by 7.2 percent (Table 6, column [3]). - Joint uncertainty between exporting AEs and importing EMDEs decreases trade by 2.4 percent (Table 6, column [5]). - Country-pair directional examples: - Exporter AE → Importer EMDE: - One SD increase of lagged uncertainty in exporting AE decreases bilateral trade with EMDE by 1.6 percent. - One SD increase of lagged uncertainty in importing EMDE decreases bilateral trade with AE by 1.5 percent (Table 6, column [4]). - Exporter EMDE → Importer AE: - One SD increase of lagged uncertainty in exporting EMDE decreases bilateral trade with AE by 4.3 percent. - (Continuation of results truncated in source excerpt.) *Source: https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024139-print-pdf.pdf* ### 5.2 percent (Table 6, column [6]). These findings suggest that the impact of uncertainty is higher when the ### wpiea2024139-print-pdf - 5.2 percent (Table 6, column [6]). These findings suggest that the impact of uncertainty is higher when the ### Effects by development level - Impact is higher when the exporter is an EMDE and the importer an AE; uncertainty from an AE partner has a larger effect on EMDE exporters, reflecting production and financing problems and low demand faced by firms in EMDEs. - Uncertainty in the exporting country has a negative and significant impact at the 1 percent level on bilateral EMDEs trade; uncertainty in the importing country is not significant. - A one SD increase of lagged uncertainty in the exporting country decreases intra-EMDEs bilateral trade by 4.5 percent. - For intra-AEs bilateral trade: - A one SD increase of lagged uncertainty in the importing country decreases intra-AEs bilateral trade by 2.5 percent. - A one SD increase of lagged uncertainty in the exporting country (AE) decreases bilateral trade by 1 percent. - Similar estimation using income group classification yields the same result: uncertainty on both sides impacts bilateral trade between high-income countries and medium-income countries. - Table 6 selected coefficients (beta coefficients between brackets; robust standard errors in parentheses): - Joint Uncertainty푡−1 (Unc푖,푡−1+Unc푗,푡−1): - Column [1]: -0.121*** (0.0302) [-0.0265] - Column [2]: -0.109*** (0.0265) [-0.0237] - Column [3]: -0.328*** (0.0477) [-0.0716] - Column [4]: -0.149*** (0.0286) [-0.0326] - Uncertainty from exporter푡−1 (Unc푖,푡−1): - Column [1]: -0.0703* (0.0419) [-0.00996] - Column [2]: -0.115** (0.0459) [-0.0162] - Column [3]: -0.303*** (0.0581) [-0.0430] - Column [4]: -0.318*** (0.0414) [-0.0450] - Uncertainty from importer푡−1 (Unc푗,푡−1): - Column [1]: -0.173*** (0.0415) [-0.0245] - Column [2]: -0.105*** (0.0325) [-0.0149] - Column [3]: -0.367*** (0.0775) [-0.0519] - Column [4]: 0.0159 (0.0401) [0.00225] - Observations and R2 (selected): - Observations: 32,718; 32,718; 113,448; 113,448; 79,887; 79,887; 193,402; 193,402 - R2: 0.967; 0.967; 0.900; 0.900; 0.842; 0.842; 0.798; 0.798 - Fixed effects: Exporter Fixed Effects, Pair Fixed Effects, Year Fixed Effects included. ### Intra versus extra-continental trade - Intraregional (intra-) results: - Intra-African and intra-European trade are most negatively affected by uncertainty. - Uncertainty from exporting countries negatively and significantly impacts intraregional trade for Africa and Europe at the 1 percent level. - A one SD increase of uncertainty in exporting countries: - Decreases intra-African bilateral trade by 4.2 percent. - Decreases intra-European bilateral trade by 2.1 percent. - For intra-European trade, a one SD increase of uncertainty in the importing country decreases trade by 1.2 percent (Table 7, columns [2] and [8]). - Extra-continental results: - Table A4 results: about half of the coefficients in front of lagged joint uncertainty are negative and significant at the 1 percent level. - In most cases, a surge in lagged joint uncertainty decreases extra-continental bilateral trade. - For Africa: - A one SD hike in lagged joint uncertainty decreases bilateral trade between exporting African countries and the rest of the world (ROW). - A hike in lagged joint uncertainty between an importing African country and the ROW does not significantly impact their bilateral trade flows. - This suggests uncertainty does not significantly hit African countries' imports from the ROW, potentially reflecting dependency on imports for basic needs. - Table 7 selected coefficients (intracontinental trade; beta coefficients in brackets; robust standard errors in parentheses): - Joint Uncertainty푡−1: - Africa: -0.114 (0.0750) [-0.0248] - `America: -0.0361 (0.0596) [-0.00787] - Asia: 0.0908 (0.0720) [0.0198] - Europe: -0.115*** (0.0325) [-0.0251] - Pacific: 0.354 (0.331) [0.0772] - Uncertainty from exporter푡−1: - Africa: -0.295*** (0.0985) [-0.0418] - `America: -0.0510 (0.0861) [-0.0072] - Asia: 0.248** (0.101) [0.0351] - Europe: -0.145*** (0.0466) [-0.0206] - Pacific: 0.00690 (0.227) [0.0010] - Uncertainty from importer푡−1: - Africa: 0.0640 (0.111) [0.00906] - `America: -0.0226 (0.0784) [-0.0032] - Asia: -0.0571 (0.101) [-0.0081] - Europe: -0.0851* (0.0451) [-0.0121] - Pacific: 0.643 (0.474) [0.0910] - Observations and R2 (selected): - Observations: 28,242; 28,242; 16,381; 16,381; 33,521; 33,521; 32,243; 32,243; 198; 198 - R2: 0.750; 0.751; 0.896; 0.896; 0.880; 0.880; 0.952; 0.952; 0.979; 0.979 ### Conditional factors — overview - The study integrates multiplicative interaction terms to capture non-linearities: trade integration (trade intensity, GVC participation), geopolitical risk, and bilateral geopolitical distance (diplomatic disagreement). - Equation (8) specification includes interaction terms 푀푖푗푡×Unc푖푡, 푀푖푗푡×Unc푗푡 and controls (gravity variables, exporter/importer fixed effects, pair and year fixed effects). ### Horizontal integration (trade intensity) - Trade intensity measures: TI1 and TI2 as defined (bilateral trade divided by sum of total trade or sum of outputs); references to Frankel and Rose (1997, 1998), Baxter and Kouparitsas (2005), Tapsoba (2009). - Key marginal effects (percentiles of trade intensity): - A one SD increase in lagged joint uncertainty: - Associated with an approximately 7.3 percent decrease in bilateral trade between countries at the 25th percentile of trade intensity. - Decreases bilateral trade by 0.37 percent between countries at the 70th percentile of trade intensity. - The impact becomes positive at the 75th percentile of trade intensity. - Uncertainty in the exporting country: - A one SD hike leads to approximately 6.8 percent decrease at the 25th percentile of trade intensity. - A one SD increase leads to a 1.1 percent decrease at the 75th percentile; impact becomes positive at the 85th percentile. - Uncertainty in the importing country: - A one SD increase leads to a 2.8 percent decrease at the 25th percentile. - Leads to a 1.5 percent increase at the 75th percentile. - Interpretation: negative impacts of uncertainty are mitigated by higher trade intensity; deeply integrated partners become "unavoidable" trading partners and are less affected. ### Vertical integration (GVC participation) - GVCs account for almost half of global trade (World Bank, 2020); about 70 percent of international trade involves GVCs; 15 percent of firms are involved and capture 80 percent of total trade (World Bank, 2020). - Findings on marginal effects by joint GVC participation percentiles: - A one SD increase in lagged joint uncertainty: - Decreases bilateral trade by 2.6 percent for 25 percent joint GVCs participation (less integrated countries). - Decreases bilateral trade by 5.6 percent for 75 percent joint GVCs participation (well-integrated countries). - Uncertainty in the exporting country: - One SD increase leads to approximately 2 percent decrease in bilateral trade for exporting country GVC participation at 25 percent. - One SD increase leads to 6 percent decrease for exporting country GVC participation at 75 percent. - Uncertainty in the importing country: - One SD hike leads to approximately 0.1 percent decrease for importing country GVC participation at 25 percent. - One SD hike leads to 2.5 percent decrease for importing country GVC participation at 75 percent. - Interpretation: higher involvement in GVCs—especially for exporting countries—amplifies the negative impact of uncertainty on bilateral trade. ### Geopolitical risk - Joint geopolitical risk between country pairs ranges between 0 and 6.63; geopolitical risk in exporting or importing countries ranges between 0 and 4.35. - Marginal effects by geopolitical risk levels: - A one SD hike in lagged joint uncertainty: - Decreases bilateral trade by 1.9 percent for partners with low joint geopolitical risk (level of geopolitical risk of 1). - Decreases bilateral trade by 9.6 percent for partners with high joint geopolitical risk (level of geopolitical risk of 3). - Uncertainty in the exporting country: - For geopolitical risk of zero in the exporting country, a one SD hike in lagged uncertainty leads to an approximately 0.3 percent increase in bilateral trade. - For geopolitical risk of one in the exporting country, a one SD increase in uncertainty leads to a 6.8 percent decrease in bilateral trade. - For geopolitical risk of two in the exporting country, a one SD increase leads to a 14 percent decrease in bilateral trade. - Interpretation: higher geopolitical risk deepens the negative impact of uncertainty for both exporters and importers. ### Diplomatic disagreement (bilateral geopolitical distance) - Diplomatic disagreement deepens the impact of uncertainty. - Marginal effects by diplomatic disagreement percentiles: - A one SD increase in lagged joint uncertainty: - Leads to an approximately 0.067 percent decrease in bilateral trade at the 25th percentile of diplomatic disagreement (friendly countries). - Leads to a 6.4 percent decrease in bilateral trade at the 75th percentile of diplomatic disagreement (high diplomatic rivals). - For uncertainty in the exporting country: - A one SD hike in lagged uncertainty is associated with a 1 percent decrease in bilateral trade between countries at the 25th percentile of diplomatic disagreement (additional higher-percentile impacts are reported in the main estimates). - Interpretation: diplomatic rivalry amplifies how uncertainty reduces bilateral trade, consistent with prior findings (Jakubik and Ruta, 2023) when allowing country-level variation in uncertainty. *IMF Working Paper extract: The Heterogeneous Effects of Uncertainty on Trade (selected sections and tables).* ### 5.6 percent decrease in bilateral trade between countries at the 75 ### 5.6 percent decrease in bilateral trade between countries at the 75th percentile. Unlike the exporting country, in the importing country, a one SD hike in lagged uncertainty is associated with an increase in bilateral trade between friendly countries by 1 percent, against a 2.8 percent decrease in bilateral trade between nonfriendly countries. Similarly, to geopolitical risk in trading partners’ countries, the geopolitical distance between bilateral partners aggravates the impact of uncertainty on bilateral trade. ### Main empirical findings - A one SD increase in lagged joint uncertainty (baseline specification) is associated with a decline in bilateral trade by 4.5 percent (baseline reference). - At the 75th percentile, bilateral trade decreases by 5.6 percent. - In the importing country, a one SD hike in lagged uncertainty: - is associated with an increase in bilateral trade between friendly countries by 1 percent, - is associated with a decrease in bilateral trade between nonfriendly countries by 2.8 percent. - Geopolitical distance between bilateral partners aggravates the negative impact of uncertainty on bilateral trade. ### Robustness: alternative measures of global uncertainty and additional controls - Using lagged average uncertainty ( (Unc_i,t-1 + Unc_j,t-1)/2 ), a one SD increase is associated with a decline in bilateral trade by 3.9 percent (Table 8, column [1]). - Using lagged maximal uncertainty (Max(Unc_i,t-1; Unc_j,t-1)), a one SD increase is associated with a decline in bilateral trade by 2.9 percent (Table 8, column [2]). - Using lagged minimal uncertainty (Min(Unc_i,t-1; Unc_j,t-1)), a one SD increase is associated with a decline in bilateral trade by 3.4 percent (Table 8, column [3]). - The difference in lagged uncertainty (Unc_i,t-1 − Unc_j,t-1) and the absolute difference |Unc_i,t-1 − Unc_j,t-1| are not statistically significant (Table 8, columns [4] and [5]). - Controlling for additional variables (volatility of credit in exporting and importing countries, terms of trade in exporting and importing countries, industrial value added on both sides, and bilateral exchange rate) confirms baseline estimates: - A 1 SD hike in lagged joint uncertainty reduces trade by 3.1 percent (Table A9, column [1]). - An increase of lagged uncertainty in the importing country by one SD is associated with a decline in bilateral trade by 2.24 percent (Table A9, column [2]). - An increase of lagged uncertainty in the exporting country by one SD is associated with a decline in bilateral trade by 1.79 percent (Table A9, column [2]). - These robustness checks indicate demand effects are more important than supply effects. ### Other types and sources of uncertainty (trade-related and pandemic-related) - World Trade Uncertainty Index (Ahir, Bloom, and Furceri (2018)) shows the highest hike in trade uncertainty during the US-China trade tensions (Figure 8). - World Pandemic Uncertainty Index (Ahir, Bloom, and Furceri (2022)) records highest episodes during SARS (2002–2003) and COVID-19 (2020–23) (Figure 9). - Trade uncertainty and pandemic uncertainty both negatively and significantly impact bilateral trade: - An increase of lagged joint trade uncertainty by one SD would lead to a reduction in bilateral trade of 1.6 percent (Table 9, column [1]). - An increase of lagged joint pandemic uncertainty by one SD would lead to a reduction in bilateral trade of 4.6 percent (Table 10, column [1]). - Trade uncertainty by source: - A one SD increase in lagged trade uncertainty from the exporting country reduces bilateral trade by 1.3 percent (Table 9, column [2]). - A one SD increase in lagged trade uncertainty from the importing country reduces bilateral trade by 0.7 percent (Table 9, column [2]). - Interpretation: trade uncertainty shows a higher supply effect. - Pandemic uncertainty by source: - A one SD increase in lagged pandemic uncertainty from the importing country reduces bilateral trade by 2.6 percent (Table 10, column [2]). - A one SD increase in lagged pandemic uncertainty from the exporting country reduces bilateral trade by 2.5 percent (Table 10, column [2]). - Interpretation: pandemic-related demand effects are slightly more important than supply effects. ### Key statistics and estimation details (from tables) - Table 8 (Mean, Max, Min, and difference): - Mean_Uncertainty_t-1 coefficient: -0.353*** (Robust standard error 0.0619) [beta: -0.0386]; Observations: 426,578; R2: 0.9735. - Maximum Uncertainty_t-1 coefficient: -0.180*** (Robust standard error 0.0418) [beta: -0.0285]; Observations: 426,578; R2: 0.9734. - Minimum Uncertainty_t-1 coefficient: -0.403*** (Robust standard error 0.0641) [beta: -0.0343]; Observations: 426,578; R2: 0.9735. - Difference Uncertainty_t-1 coefficient: 0.00728 (Robust standard error 0.0341) [beta: 0.00134]; Observations: 425,993; R2: 0.9733. - Absolute Difference Uncertainty_t-1 coefficient: -0.0384 (Robust standard error 0.0418) [beta: -0.00501]; Observations: 425,993; R2: 0.9733. - Fixed effects: Exporter, Importer, Pair, Time included. - Table 9 (Trade uncertainty): - Joint Trade Uncertainty_t-1 coefficient: -0.00539*** (Robust standard error 0.000929) [beta: -0.0156]; Observations: 313,640; R2: 0.9815. - Trade Uncertainty from exporter_t-1 coefficient: -0.00723*** (Robust standard error 0.00145) [beta: -0.0134]. - Trade Uncertainty from importer_t-1 coefficient: -0.00379*** (Robust standard error 0.00120) [beta: -0.00703]. - Fixed effects: Exporter, Importer, Pair, Time included. - Table 10 (Pandemic uncertainty): - Joint Pandemic Uncertainty_t-1 coefficient: -0.00502*** (Robust standard error 0.00162) [beta: -0.0456]; Observations: 313,640; R2: 0.9815. - Pandemic Uncertainty from exporter_t-1 coefficient: -0.00492** (Robust standard error 0.00206) [beta: -0.0251]. - Pandemic Uncertainty from importer_t-1 coefficient: -0.00510** (Robust standard error 0.00251) [beta: -0.0261]. - Fixed effects: Exporter, Importer, Pair, Time included. *IMF Working Paper — The Heterogeneous Effects of Uncertainty on Trade* ### 3. GMM methodology ### 3. GMM methodology ### GMM approach and identification - The study followed a system-GMM approach (Blundell and Bond 1998) to check robustness and control endogeneity. - Acknowledged limitation: the GMM approach can be subject to weak instruments. - Instruments used: - Period lag of uncertainty as instrument for uncertainty. - Second-period lag of bilateral trade (the second-order lagged logarithm of trade) as instrument for the lagged dependent variable. - For the System GMM estimation, the period goes from 1980 to 2021, with six constructed periods; each variable for a given period represents the average on seven consecutive years. ### Main GMM estimates (summary of Table 11) - IV/GMM estimation (columns [1] and [2]) and System GMM estimation (columns [3] and [4]) both find uncertainty measures negative and statistically significant at the 1 percent level for joint uncertainty and for disaggregated exporter/importer uncertainty in many specifications. - Key coefficient estimates and associated precision (as reported in Table 11): - Joint Uncertainty (Unc_i,t + Unc_j,t) - Column [1] (IV/GMM): -0.104*** (0.0230) [ -0.0227 ] - Column [3] (System GMM): -0.211*** (0.0753) [ -0.0477 ] - Uncertainty from exporter (Unc_i,t) - Column [1] (IV/GMM): -0.145*** (0.0330) [ -0.0205 ] - Column [3] (System GMM): -0.194* (0.108) [ -0.0293 ] - Uncertainty from importer (Unc_j,t) - Column [1] (IV/GMM): -0.0644** (0.0324) [ -0.00912 ] - Column [3] (System GMM): -0.213** (0.104) [ -0.0312 ] - Lagged Log of Trade (Log of Trade_t−1) - Columns [1] and [2]: 0.772*** (0.00515) - Columns [3] and [4]: 0.577*** (0.0144) - Observations and panel dimensions reported: - Columns [1] and [2]: Observations 354,443; Number of pairs 15,354. - Columns [3] and [4]: Observations 60,664; Number of pairs 16,498. - Diagnostics and fixed effects: - Inst Count < Groups: Yes for all columns. - Over identification / Hansen: columns [3]/[4] report 135.9 and 139.5 respectively. - AR1 p-value (system GMM): 0.000 and 0.000. - AR2 p-value (system GMM): 0.167 and 0.167. - Exporter Fixed Effects: Yes in all columns. - Importer Fixed Effects: Yes in all columns. - Pair Fixed Effects: Yes in columns [1] and [2]; No in columns [3] and [4]. - Time Fixed Effects: Yes in all columns. - Note: Gravity model’s basic control variables are included. Robust standard errors are in parentheses. Significance markers: *** p<0.01, ** p<0.05, * p<0.1. ### Magnitude interpretation reported in text - The GMM findings (as summarized in the main text referencing Table 11): - A one SD increase in joint uncertainty reduces bilateral trade by 2.3 (Table 11, column [1]). - A one SD increase in uncertainty in the exporting country is associated with a decline in bilateral trade of 2.1 percent. - A one SD increase in uncertainty in the importing country is associated with a decline in bilateral trade of 1 percent. - The system GMM results (Table 11, columns [3] and [4]) yield higher coefficients, consistent with baseline results. ### Conclusion points and policy-relevant findings (from the same chapter) - Overall conclusions using an augmented gravity model with 143 countries from 1980 to 2021: - Uncertainty negatively impacts bilateral trade between countries. - The negative impact channels through both supply and demand effects: uncertainty in the exporting country and uncertainty in the importing country negatively affect bilateral trade, with a higher impact of uncertainty from the importing country. - Trade in manufacturing and mineral fuel products is more at risk. - Heterogeneity in effects: - EMDEs and non-natural resource-rich countries are more at risk. - EMDEs are more likely to be impacted in trade relations with both AEs and EMDEs. - Intra-African and intra-European trade found vulnerable to uncertainty shocks. - Non-linearities documented: - The negative impact of uncertainty is moderated by the level of trade intensity between trading partners. - The negative impact is aggravated by geopolitical risk, geopolitical distance, and the level of GVCs participation. - Policy implications highlighted: - Policymakers, especially in EMDEs, need to consider uncertainty as a key element in strategy. - Despite benefits of international trade, it is sensitive to high uncertainty in exporting or importing countries; economies integrated into GVCs can become quickly vulnerable to uncertainty. - Promoting openness and trade integration should be complemented by measures and protocols to prevent economic losses from potential uncertainty shocks. - Trading more with a partner can help counter the negative impact of uncertainty because highly integrated partners may become unavoidable trade partners. *Source: 3. GMM methodology (from wpiea2024139-print-pdf).* ### References ### References ### Foundational theoretical and empirical contributions - Abel, A. B. (1983). Optimal investment under uncertainty. American Economic Review, 73(1), 228–233. https://www.jstor.org/stable/1803942 - Hartman, R. (1972). The effects of price and cost uncertainty on investment. Journal of Economic Theory, 5(2), 258–266. https://doi.org/10.1016/0022-0531(72)90105-6 - Bernanke, B. S. (1983). Irreversibility, uncertainty, and cyclical investment. The quarterly journal of economics, 98(1), 85-106. https://doi.org/10.2307/1885568 - Oi, W. Y. (1961). The Desirability of Price Instability Under Perfect Competition. Econometrica, 29(1), 58–64. https://doi.org/10.2307/1907687 - Bar-Ilan, A., & Strange, W.C. (1996). Investment lags. The American Economic Review, 86(3), 610–622. - Blundell, R., & Bond, S. (1998). Initial conditions and moment restrictions in dynamic panel data models. Journal of Econometrics 87: 115–143. https://doi.org/10.1016/S0304-4076(98)00009-8 ### Uncertainty measures, geopolitical risk, and political economy of trade - Ahir, H., Bloom, N., & Furceri, D. (2018). The World Uncertainty Index (SSRN Scholarly Paper ID 3275033). Social Science Research Network. https://doi.org/10.2139/ssrn.3275033 - Ahir, H., Bloom, N., & Furceri, D. (2022). The World Uncertainty Index (Working Paper No. 29763; Working Paper Series). National Bureau of Economic Research. https://doi.org/10.3386/w29763 - Caldara, D., & Iacoviello, M. (2022). Measuring geopolitical risk. American Economic Review, 112(4), 1194-1225. https://www.aeaweb.org/articles?id=10.1257/aer.20191823 - Scheffel, E. M. (2016). Accounting for the political uncertainty factor. Journal of Applied Econometrics, 31(6), 1048-1064. https://doi.org/10.1002/jae.2455 - Pástor, Ľ., & Veronesi, P. (2021). Inequality aversion, populism, and the backlash against globalization. The Journal of Finance, 76(6), 2857-2906. https://doi.org/10.1111/jofi.13081 - Rodrik, D. (2018). Populism and the economics of globalization. Journal of International Business Policy, pp. 1, 12–33. https://link.springer.com/article/10.1057/s42214-018-0001-4 - Autor, D., Dorn, D., Hanson, G., & Majlesi, K. (2020). Importing political polarization? The electoral consequences of rising trade exposure. American Economic Review, 110(10), 3139-3183. https://www.aeaweb.org/articles?id=10.1257/aer.20170011 ### Uncertainty and trade: theory, evidence, and channels - Handley, K. (2014). Exporting under trade policy uncertainty: Theory and evidence. Journal of International Economics, 94(1), 50–66. https://doi.org/10.1016/j.jinteco.2014.05.005 - Handley, K., and Limão, N. (2015). Trade and investment under policy uncertainty: Theory and firm evidence. American Economic Journal: Economic Policy, 7(4), 189–222. https://doi.org/10.1257/pol.20140068 - Taglioni, D., & Zavacka, V. (2013). Innocent Bystanders: How Foreign Uncertainty Shocks Harm Exporters (SSRN Scholarly Paper ID 2240185). Social Science Research Network. https://doi.org/10.2139/ssrn.2240185 - Baley, I., Veldkamp, L., & Waugh, M. (2020). Can global uncertainty promote international trade? Journal of International Economics, 126, 103347. https://doi.org/10.1016/j.jinteco.2020.103347 - Novy, D., & Taylor, A. M. (2020). Trade and uncertainty. Review of Economics and Statistics, 102(4), 749-765. https://doi.org/10.1162/rest_a_00885 - Grier, K. B., & Smallwood, A. D. (2007). Uncertainty and Export Performance: Evidence from 18 Countries. Journal of Money, Credit and Banking, 39(4), 965–979. https://doi.org/10.1111/j.1538-4616.2007.00053.x - Sly, N. (2016). Global uncertainty and U.S. exports. Economic Review (Kansas City), 101(2), 5–25. https://ideas.repec.org/a/fip/fedker/00039.html ### Trade, global value chains, and structural gravity approaches - World Bank. (2020). World Development Report 2020: Trading for Development in the Age of Global Value Chains. World Bank. https://doi.org/10.1596/978-1-4648-1457-0 - Qiu, H., Shin, H. S., & Zhang, L. S. Y. (2023). Mapping the realignment of global value chains (No. 78). Bank for International Settlements. - Constantinescu, C., Mattoo, A., & Ruta, M. (2020). Policy uncertainty, trade, and global value chains: some facts, many questions. Review of Industrial Organization, pp. 57, 285–308. https://link.springer.com/article/10.1007/s11151-020-09772-0 - Ponte, S., & Gibbon, P. (2005). Quality standards, conventions, and the governance of global value chains. Economy and society, 34(1), 1–31. https://doi.org/10.1080/0308514042000329315 - Silva, J. S., & Tenreyro, S. (2006). The log of gravity. The Review of Economics and Statistics, 88(4), 641–658. https://doi.org/10.1162/rest.88.4.641 - Anderson, J. E., & Van Wincoop, E. (2003). Gravity with gravitas: A solution to the border puzzle. American Economic Review, 93(1), 170–192. https://www.aeaweb.org/articles?id=10.1257/000282803321455214 - Yotov, Y. V., Piermartini, R., & Larch, M. (2016). An advanced guide to trade policy analysis: The structural gravity model. WTO iLibrary. https://www.wto.org/english/res_e/booksp_e/advancedwtounctad2016_e.pdf ### Business cycles, synchronization, and trade intensity - Baxter, M., & Kouparitsas, M. A. (2005). Determinants of business cycle comovement: a robust analysis. Journal of Monetary Economics, 52(1), 113–157. https://doi.org/10.1016/j.jmoneco.2004.08.002 - Frankel, J. A., & Rose, A. K. (1997). Is EMU more justifiable ex post than ex ante? European Economic Review, 41(3-5), 753-760. https://doi.org/10.1016/S0014-2921(97)00034-2 - Frankel, J. A., & Rose, A. K. (1998). The endogeneity of the optimum currency area criteria. The Economic Journal, 108(449), 1009-1025. https://doi.org/10.1111/1468-0297.00327 - Calderon, C., Chong, A., & Stein, E. (2007). Trade intensity and business cycle synchronization: Are developing countries any different? Journal of International Economics, 71(1), 2-21. https://doi.org/10.1016/j.jinteco.2006.06.001 - Inklaar, R., Jong-A-Pin, R., and De Haan, J. (2008). Trade and business cycle synchronization in OECD countries—A re-examination. European Economic Review, 52(4), 646-666. https://doi.org/10.1016/j.euroecorev.2007.05.003 - Tapsoba, S. J. A. (2009). Trade intensity and business cycle synchronicity in Africa. Journal of African Economies, 18(2), 287-318. https://academic.oup.com/jae/article/18/2/287/729797 ### Recent working papers, country studies, and applied analyses - Jakubik, A., and Ruta, M. (2023). Trading with Friends in Uncertain Times (No. 2023/124). International Monetary Fund. https://www.imf.org/en/Publications/WP/Issues/2023/06/16/Trading-with-Friends-in-Uncertain-Times-534717 - Graziano, A., Handley, K., and Limão, N. (2018). Brexit Uncertainty and Trade Disintegration (Working Paper No. 25334; Working Paper Series). National Bureau of Economic Research. https://doi.org/10.3386/w25334 - Matzner, A., Meyer, B., & Oberhofer, H. (2023). Trade in Times of Uncertainty (2023). CESifo Working Paper No. 10284. Available at SSRN: https://ssrn.com/abstract=4368213 - Novy, D., & Taylor, A. M. (2020). Trade and uncertainty. Review of Economics and Statistics, 102(4), 749-765. https://doi.org/10.1162/rest_a_00885 - Nana, I., & Ouedraogo, I. (2023). Friendship is as Important as Neighborhood: The impact of Geopolitical Distance on Bilateral Trade. 2023. Working Paper - CERDI ⟨hal-04266229⟩ https://hal.science/hal-04266229/document - Nana, I., Motelle, S. I., & Starnes, S. K. (2023). The Trade-Growth Nexus: Evidence of Causality from Innovative Instruments for Trade. Policy Research working paper; no. WPS 10645 Washington, D.C.: World Bank Group. http://documents.worldbank.org/curated/en/099550412132341350/IDU0a68116e90579804937091a70918736db8633 ### Case studies, policy commentary, and sectoral evidence - Freund, C., Mattoo, A., Mulabdic, A., & Ruta, M. (2023). US-China decoupling: Rhetoric and reality. VoxEU Column, 31. - Pangestu, M.E. & Trotsenburg, A.V. (2022). Trade restrictions are inflaming the world food crisis in a decade. World Bank. (Accessed on December 10, 2023) https://blogs.worldbank.org/voices/trade-restrictions-are-inflaming-worst-food-crisis-decade - World Bank (2023). Conflict in the Middle East Could Bring ‘Dual Shock’ to Global Commodity Markets. Press Release October 30, 2023. NO: 2024/025/DEC - Bonnell, C., & McHugh, D. (2024). “How attacks on ships in the Red Sea by Yemen’s Houthi rebels are crimping global trade” Yahoo News January 12, 2024. https://news.yahoo.com/houthi-attacks-ships-red-sea-093757343.html - Raulatu, B. L., Ugbem, O. V., Augustine, U., & Paul, F. M. (2019). Uncertainties in Global Economic Policy and Nigeria’s Export Earnings. International Journal of Business, Economics and Management, 6(1), 23–38. https://doi.org/10.18488/journal.62.2019.61.23.38 *Source: wpiea2024139-print-pdf - References* --- _Source: https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024139-print-pdf.pdf_