## Introduction

## Source details

**Canonical URL:** [Introduction](https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025180-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2025/english/wpiea2025180-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2025/english/wpiea2025180-source-pdf.pdf.json)

---

### Context and high-level findings
- Since NAFTA in 1992, Mexico steadily increased its share of U.S. imports and became the most important U.S. trading partner in 2023, representing 15.4 percent of total U.S. imports.
- Mexico’s share increase coincided with a decline in China’s share from its peak in 2017.
- Global trade growth was interrupted by multiple shocks (GFC, Arab Spring, Brexit) and more recently by geopolitical fragmentation associated with U.S.–China trade tensions, the pandemic, and Russia’s invasion of Ukraine.
- Geopolitical fragmentation increased trade within political blocks and among non-aligned countries, driving relocation of global value chains (GVCs) for diversification and logistical resilience.

### Mechanisms documented
- The 2018–19 U.S. tariffs on Chinese products led to:
  - A reduction in U.S. imports from China.
  - An increase in U.S. imports from countries producing substitutes for Chinese exports (e.g., Thailand and Mexico).
  - Evidence that firms in Mexico integrated in GVCs increased their imports of inputs from China and other Asian countries (Thailand, Vietnam, Japan, Korea, Indonesia).
- Relocation of GVCs responds to cost competitiveness, productivity, proximity to the U.S., and established logistics networks.

### Trade developments and tariff chronology
- Between 2017–2023:
  - The share of imports from China in total U.S. imports declined by almost 8 percentage points.
  - Mexico, the Euro Area, and Vietnam each increased penetration of the U.S. market by around 2 percentage points.
- U.S. tariff chronology and sizing:
  - 2018: tariffs on imports from China amounting to about US$50 billion.
  - September 2018: additional tariffs on US$200 billion worth of Chinese products, initially set at 10 percent.
  - January 2019: those tariffs were later increased to 25 percent.
  - By the end of 2019, most goods traded between the U.S. and China were subject to tariffs; these have been mostly maintained as of end 2024.
- Sectoral winners in U.S. imports (2017–2023):
  - Mexico’s exports of electronics and semiconductors increased their share of U.S. imports in the sector by 6 percentage points.
  - Transportation equipment (including automobiles and trucks) increased its share in U.S. imports by 6.5 percentage points.

### Empirical strategy and key results (trade)
- Data and method:
  - Monthly six-digit product-level trade data from UNComtrade.
  - Difference-in-differences design exploiting variation in the 2018–19 U.S. tariffs on Chinese products.
  - Specification includes product-level tariff change interacted with a post-July 2018 indicator, lagged dependent variable, and product and time fixed effects.
  - Coefficient of interest (β) represents the elasticity (percent response) of U.S. imports to a one-percent tariff change.
- Country-level difference-in-differences estimates (Tariff increase × Post Jul 2018; SEs clustered by HS6-digit; significance: *** p<0.01, ** p<0.05, * p<0.1):
  - CHN: -1.285*** (SE 0.076)
  - MEX: 0.364*** (SE 0.107)
  - CAN: 0.084 (SE 0.090)
  - GTM: 1.108*** (SE 0.326)
  - HND: 1.025*** (SE 0.373)
  - IDN: 0.749*** (SE 0.195)
  - MYS: 1.050*** (SE 0.253)
  - VNM: 0.415** (SE 0.166)
  - THA: 0.642*** (SE 0.154)
  - AUT: 0.440** (SE 0.183)
  - POL: 0.796*** (SE 0.215)
- Lagged ln(US imports) (examples):
  - CHN: 0.414*** (0.007)
  - MEX: 0.539*** (0.009)
  - CAN: 0.419*** (0.008)
- Selected model fit and sample sizes:
  - CHN: N = 328,669; R-squared = 0.903
  - MEX: N = 234,849; R-squared = 0.923
  - CAN: N = 289,195; R-squared = 0.914
- Estimated USD impacts (simple calculation based on estimated elasticities):
  - Mexico: initial impact around US$40 billion; total estimated impact about US$70 billion (the US$70 billion accounts for around 45 percent of the total increase in Mexico’s exports to the U.S. during 2017–2023).
  - China: tariffs are estimated to have reduced Chinese exports to the U.S. by close to US$200 billion at the peak.
  - Canada: estimated positive impact about US$30 billion.
  - Malaysia: estimated positive impact about US$20 billion.
- Dynamics:
  - A divergence emerged in 2018–19 between U.S. imports from China targeted by the tariffs (treated) and those not affected (control); the divergence widened after the pandemic and since Russia’s invasion of Ukraine.
  - For Mexico, the gap between exports affected by the tariff and those not covered increased over time, indicating firms in GVCs adjusted to a persistent shift.

### Mexico’s sectoral gains and GVC participation
- Aggregate change:
  - Mexico’s exports to the U.S. increased by US$163 billion between 2017–2023.
  - Sectoral composition of that increase:
    - Vehicles and cars: about 30 percent of the total increase.
    - Computing and electronic products: about 15 percent of the total increase.
  - Most of the increase came from manufactures concentrated in sectors with established U.S. linkages and rising U.S. demand.
- Mexico broadening its participation in GVCs:
  - Mexico’s imports of products affected by the 2018–19 U.S. tariffs on China have increased, although by significantly smaller amounts than the rise of Mexico’s exports to the U.S.
  - Many of those import increases originate from other emerging markets including in Asia.

### Estimated impact of 2018 U.S. tariffs on Mexico’s imports (country-level and goods-type)
- Aggregate estimated impacts (2017–2023):
  - Increase of Mexico’s imports from China by about US$14 billion associated with the U.S. tariffs and subsequent global events.
  - Increases in Mexico’s imports linked to U.S. trade policy:
    - Korea by close to US$6 billion.
    - Vietnam by close to US$1.5 billion.
- Table 2 select coefficient (Dependent variable: ln(MEX imports), Tariff increase × Post Jul 2018):
  - CHN: 0.361*** (SE 0.097)
  - USA: 0.104 (SE 0.090)
  - VNM: 1.156*** (SE 0.349)
- Composition shifts:
  - Mexico’s recent import increases concentrated in intermediate goods, consistent with deepening participation in GVCs.
  - Mexico’s imports of intermediate goods from the U.S. increased the most between 2017-2023 compared with other origins.
  - Trading partners explaining most of the increase in Mexico’s imports: China, Vietnam, Malaysia, and Germany; in all cases most of the increase corresponds to intermediate goods feeding supply chains.
- Bilateral flows:
  - Mexico’s exports to the U.S. increased by more than US$160 billion between 2017-2023.
  - Mexico’s imports from China increased by only US$40 billion over the same period.
  - Intermediate goods explain about US$25 billion of the rise in Mexico’s imports from China.
- Changes in goods categories (2017-2023):
  - Intermediate goods increased by close to 40 percent.
  - Consumption goods increased by about 100 percent.
  - Investment (capital) goods increased by about 70 percent.
  - Mexico’s imports of final goods (consumption and capital) from China increased as a share of total imports by about close to 6 percentage points.
- Product-level patterns:
  - Large increases in consumption goods from China, including cars and smartphones.
  - Chinese vehicles gained important share in the Mexican market, particularly from domestically produced vehicles.
  - Product sets with largest increases in Mexico’s imports from China (e.g., home electronics, small vehicles, smartphones) differ from those with largest increases in Mexico’s exports to the U.S. (e.g., trucks, large cars, semiconductors).
  - Small vehicle (1000-1500 cc) imports from China appear associated with domestic consumption.

### FDI dynamics and empirical results (industry-level)
- FDI flows:
  - Mexico’s share in global FDI to EMs rose from 6 percent on average in the 2010s to close to 10 percent in 2023.
  - FDI composition by origin:
    - U.S.: around 40 percent of total FDI to Mexico.
    - Euro Area: 20 percent of total.
    - China: less than 2 percent of total FDI to Mexico and less than 0.02 percent of Mexico’s GDP.
  - Geographic and sectoral concentration:
    - FDI inflows clustered in northern states (Sonora, Nuevo Leon, Jalisco, Chihuahua) and Mexico City.
    - Sectors receiving most FDI: manufacturing (particularly vehicles and electronics), finance, and real estate.
- Empirical approach:
  - Difference-in-difference on four-digit industry FDI data (state by INEGI 4-digit industry, quarterly frequency).
  - Regressors: ΔOutputTariff × Post Jul 2018, ΔUpstreamTariff × Post Jul 2018, ΔDownstreamTariff × Post Jul 2018, lagged ln(FDI), state, industry, and time fixed effects.
  - Input-output linkages computed using INEGI input-output table; mapping via NAICS 4-digit to HS6-digit.
  - Sample period: 2016Q1-2023Q4. SEs clustered by state.
- Selected FDI regression results (Dependent variable: ln(FDI), N. of observations 50,126, R-squared 0.953; significance: *** p<0.01, ** p<0.05, * p<0.1):
  - Post Jul 2018 × Output Tariff: 0.876*** (0.130) in column (1); 0.831*** (0.133) in column (2).
  - Post Jul 2018 × Upstream Tariff: 1.160*** (0.393) in column (2); -0.349 (0.419) in column (3).
  - Post Jul 2018 × Downstream Tariff: 0.998*** (0.182) in column (3); 0.664*** (0.170) in column (4).
  - Lagged ln(FDI): ~0.822*** (0.015) across specifications.
- Manufacturing vs non-manufacturing (Dependent variable: ln(FDI), N. of observations 45,678, R-squared 0.953):
  - Manufacturing × Post Jul 2018 × Output Tariff: 0.269** (0.109) in column (1); 0.485** (0.188) in column (2).
  - Manufacturing × Post Jul 2018 × Upstream Tariff: 0.931** (0.447) in column (1); -0.476 (0.492) in column (2).
  - Manufacturing × Post Jul 2018 × Downstream Tariff: 0.951*** (0.254) in column (1); 1.195*** (0.393) in column (2).
  - Non-manufacturing × Post Jul 2018 × Downstream Tariff: 0.619** (0.287) in column (1); 0.570** (0.242) in column (2).
  - Manufacturing lagged ln(FDI): 0.877*** to 0.879*** (0.017). Non-manufacturing lagged ln(FDI): 0.759*** to 0.757*** (0.021-0.022).
- Interpretation:
  - FDI to Mexico increased more in industries affected by the 2018-19 U.S. tariffs on Chinese products.
  - Statistically significant downstream tariff effects suggest indirect spillovers via input-output linkages, including to non-tradable sectors.
  - Upstream tariffs show mixed/insignificant impacts in several specifications.
  - Since 2017, FDI to northern states increased more than to southern states, aligning with concentration of firms exporting to the U.S.

### Policy implications and priorities
- Mexico can capitalize on deeper integration with the North American market and ongoing reshaping of GVCs, given:
  - A developed manufacturing system.
  - Availability of human capital.
  - Cost competitiveness.
  - Proximity to the U.S. market.
- FDI has been instrumental in expanding Mexico’s productive capacity to supply U.S. demand; much of that FDI originates from the U.S.
- Recommended priorities (as outlined in the 2024 Mexico Article IV Consultation Staff Report):
  - Improving environment for private investment and business development.
  - Opening markets to private sector participation.
  - Introducing regulatory reforms.
  - Broadening access to financial services.
  - Better-targeted public investment to relieve infrastructure bottlenecks, notably in energy, transport, and water.
  - Maintaining open trade policies to underpin business investment and bolster international competitiveness.

*IMF Working Paper: Relocation of Global Value Chains: The Role of Mexico — Introduction (content unit: wpiea2025180-source-pdf - Introduction).*

### Introduction ...........................................................................................................

### Introduction

### Context and high-level findings
- Since NAFTA in 1992, Mexico has steadily increased its share of U.S. imports and became the most important U.S. trading partner in 2023, representing 15.4 percent of total U.S. imports.
- Mexico’s growing integration with the U.S. has continued despite intensified geoeconomic fragmentation; Mexico’s share increase has coincided with a decline in China’s share from its peak in 2017.
- Global trade growth has been interrupted by multiple shocks (GFC, Arab Spring, Brexit) and more recently by geopolitical fragmentation associated with U.S.–China trade tensions, the pandemic, and Russia’s invasion of Ukraine.
- Geopolitical fragmentation has increased trade within political blocks and among non-aligned countries, and has driven relocation of global value chains (GVCs) for diversification and logistical resilience.

### Mechanisms documented
- The 2018–19 U.S. tariffs on Chinese products led to:
  - A reduction in U.S. imports from China.
  - An increase in U.S. imports from countries producing substitutes for Chinese exports (e.g., Thailand and Mexico).
  - Evidence that firms in Mexico integrated in GVCs increased their imports of inputs from China and other Asian countries (Thailand, Vietnam, Japan, Korea, Indonesia).
- The relocation of GVCs responds to cost competitiveness, productivity, proximity to the U.S., and established logistics networks.

---

### Trade Developments and Data
- Between 2017–2023:
  - The share of imports from China in total U.S. imports declined by almost 8 percentage points.
  - Mexico, the Euro Area, and Vietnam each increased penetration of the U.S. market by around 2 percentage points.
- U.S. tariff chronology and sizing:
  - 2018: tariffs on imports from China amounting to about US$50 billion.
  - September 2018: additional tariffs on US$200 billion worth of Chinese products, initially set at 10 percent.
  - January 2019: those tariffs were later increased to 25 percent.
  - By the end of 2019, most goods traded between the U.S. and China were subject to tariffs; these have been mostly maintained as of end 2024.
- Sectoral winners in U.S. imports (2017–2023):
  - Mexico’s exports of electronics and semiconductors increased their share of U.S. imports in the sector by 6 percentage points.
  - Transportation equipment (including automobiles and trucks) increased its share in U.S. imports by 6.5 percentage points.

---

### Trade: Empirical Strategy and Results
- Data and method:
  - Monthly six-digit product-level trade data from UNComtrade.
  - Difference-in-differences design exploiting variation in the 2018–19 U.S. tariffs on Chinese products.
  - Specification includes product-level tariff change interacted with a post-July 2018 indicator, lagged dependent variable, and product and time fixed effects.
  - The coefficient of interest (β) represents the elasticity (percent response) of U.S. imports to a one-percent tariff change.
- Key quantitative results:
  - Difference-in-differences estimates show Mexico and other partners increased exports to the U.S. more in products affected by the 2018–19 tariffs.
  - Estimated USD impacts (simple calculation based on estimated elasticities):
    - Mexico: initial impact around US$40 billion; total estimated impact about US$70 billion (the US$70 billion accounts for around 45 percent of the total increase in Mexico’s exports to the U.S. during 2017–2023).
    - China: tariffs are estimated to have reduced Chinese exports to the U.S. by close to US$200 billion at the peak.
    - Canada: estimated positive impact about US$30 billion.
    - Malaysia: estimated positive impact about US$20 billion.
  - Table 1 (selected coefficients; sample period: 2016M1–2023M12; SEs clustered by HS6-digit; significance: *** p<0.01, ** p<0.05, * p<0.1):
    - Tariff increase × Post Jul 2018 coefficients:
      - CHN: -1.285*** (SE 0.076)
      - MEX: 0.364*** (SE 0.107)
      - CAN: 0.084 (SE 0.090)
      - GTM: 1.108*** (SE 0.326)
      - HND: 1.025*** (SE 0.373)
      - IDN: 0.749*** (SE 0.195)
      - MYS: 1.050*** (SE 0.253)
      - VNM: 0.415** (SE 0.166)
      - THA: 0.642*** (SE 0.154)
      - AUT: 0.440** (SE 0.183)
      - POL: 0.796*** (SE 0.215)
    - Lagged ln(US imports) coefficients are positive and significant across countries (examples):
      - CHN: 0.414*** (0.007)
      - MEX: 0.539*** (0.009)
      - CAN: 0.419*** (0.008)
    - Number of observations and R-squared (selected):
      - CHN: N = 328,669; R-squared = 0.903
      - MEX: N = 234,849; R-squared = 0.923
      - CAN: N = 289,195; R-squared = 0.914
- Dynamics:
  - During 2018–19 a divergence emerged between U.S. imports from China targeted by the tariffs (treated) and those not affected (control). The divergence widened after the pandemic and since Russia’s invasion of Ukraine.
  - For Mexico, the gap between exports affected by the tariff and those not covered increased over time, indicating firms involved in GVCs adjusted to a persistent shift in global trade dynamics.

---

### Mexico’s Sectoral Gains and GVC Participation
- Aggregate change:
  - Mexico’s exports to the U.S. increased by US$163 billion between 2017–2023.
  - Sectoral composition of that increase:
    - Vehicles and cars: about 30 percent of the total increase.
    - Computing and electronic products: about 15 percent of the total increase.
  - Most of the increase in Mexico’s exports to the U.S. came from manufactures concentrated in sectors with established U.S. linkages and rising U.S. demand.
- Mexico broadening its participation in GVCs:
  - Using a similar product-level difference-in-differences design, the authors estimate how U.S. tariffs on China affected Mexico’s imports from China, the U.S., and other partners.
  - Results indicate Mexico’s imports of products affected by the 2018–19 U.S. tariffs on China have increased, although by significantly smaller amounts than the rise of Mexico’s exports to the U.S., and many of those import increases originate from various other emerging markets including in Asia.

---

*IMF Working Paper: Relocation of Global Value Chains: The Role of Mexico — Introduction (content unit: wpiea2025180-source-pdf - Introduction).*

### introduction of the U.S. tariffs on imports from China, followed by global events accelerating geoeconomic

### wpiea2025180-source-pdf - introduction of the U.S. tariffs on imports from China, followed by global events accelerating geoeconomic

### Estimated impact of 2018 U.S. tariffs on Mexico’s imports
- Introduction of the U.S. tariffs on imports from China, followed by global events (pandemic, war in Ukraine), are associated with an increase of Mexico’s imports from China by about US$14 billion.
- Changes in U.S. trade policies are linked to increases in Mexico’s imports from:
  - Korea by close to US$6 billion.
  - Vietnam by close to US$1.5 billion.
- Table 2 (select coefficients, Dependent variable: ln(MEX imports)):
  - Tariff increase × Post Jul 2018 for CHN: 0.361*** (SE 0.097).
  - Tariff increase × Post Jul 2018 for USA: 0.104 (SE 0.090).
  - Tariff increase × Post Jul 2018 for VNM: 1.156*** (SE 0.349).
  - Lagged ln(MEX imports) coefficients range (examples): CHN 0.363*** (0.009), USA 0.374*** (0.009), VNM 0.418*** (0.016).
- Sample period: 2016M1-2023M12. SEs clustered by HS6-digit code. Significance: *** p<0.01, ** p<0.05, * p<0.1.
- Figure 9 (summary): Estimated impacts of the U.S. tariff on Mexico’s imports by country of origin presented in US$ billions for 2017-2023, with significance markers.

### Composition shifts and deeper participation in global value chains
- Mexico’s recent import increases are concentrated in intermediate goods, consistent with deepening participation in global value chains.
- Mexico’s imports of intermediate goods from the U.S. increased the most between 2017-2023 compared with other origins, reflecting accelerated U.S.–Mexico trade integration.
- Trading partners explaining most of the increase in Mexico’s imports: China, Vietnam, Malaysia, and Germany. In all cases most of the increase corresponds to intermediate goods feeding supply chains.
- References to supporting literature: Utar and others, 2023; Freund and others, 2024; Fajgelbaum and others, 2024; Alfaro and Chor, 2023 (as reported in source).

### Bilateral flows: Mexico’s exports to the U.S. versus imports from China
- Mexico’s exports to the U.S. increased by more than US$160 billion between 2017-2023.
- Mexico’s imports from China increased by only US$40 billion over the same period.
- Decomposition of Mexico’s imports from China:
  - Intermediate goods explain about US$25 billion of the rise.
- Figure 11 (summary): Cumulative change 2017-2023 in USD billions showing divergence: Exports to the US >> Imports from China.

### Changes in goods categories and product-level patterns
- Between 2017-2023 percentage changes in Mexico’s imports from China by goods type:
  - Intermediate goods increased by close to 40 percent.
  - Consumption goods increased by about 100 percent.
  - Investment (capital) goods increased by about 70 percent.
- Resulting change in import shares:
  - Mexico’s imports of final goods (consumption and capital) from China increased as a share of total imports by about close to 6 percentage points.
- Product-level observations:
  - Large increases in consumption goods from China, including cars and smartphones.
  - Sales data show Chinese vehicles gained an important share in the Mexican market, particularly gaining market share from domestically produced vehicles.
  - Six-digit product-level analysis: products with largest increases in Mexico’s imports from China (e.g., home electronics, small vehicles, smartphones) differ from products with largest increases in Mexico’s exports to the U.S. (e.g., trucks, large cars, semiconductors).
  - Small vehicle (1000-1500 cc) imports from China appear associated with domestic consumption.

### Foreign Direct Investment (FDI) dynamics to Mexico
- Increased U.S.–Mexico trade flows have been accompanied by robust FDI flows from the U.S. to Mexico.
- Mexico’s share in global FDI to EMs rose from 6 percent on average in the 2010s to close to 10 percent in 2023.
- FDI composition by origin (reported shares):
  - U.S.: around 40 percent of total FDI to Mexico.
  - Euro Area: 20 percent of total.
  - China: less than 2 percent of total FDI to Mexico and less than 0.02 percent of Mexico’s GDP.
- Geographic and sectoral concentration:
  - FDI inflows clustered in northern states (Sonora, Nuevo Leon, Jalisco, Chihuahua) and Mexico City.
  - Sectors receiving most FDI: manufacturing (particularly vehicles and electronics), finance, and real estate.

### Empirical strategy and FDI results (industry-level)
- Empirical approach:
  - Difference-in-difference on four-digit industry FDI data (state by INEGI 4-digit industry, quarterly frequency).
  - Regressors: ΔOutputTariff × Post Jul 2018, ΔUpstreamTariff × Post Jul 2018, ΔDownstreamTariff × Post Jul 2018, lagged ln(FDI), state, industry, and time fixed effects.
  - Input-output linkages computed using INEGI input-output table; mapping via NAICS 4-digit to HS6-digit.
  - Sample period: 2016Q1-2023Q4. SEs clustered by state.
- Table 3 (select coefficients, Dependent variable: ln(FDI), N. of observations 50,126, R-squared 0.953):
  - Post Jul 2018 × Output Tariff: 0.876*** (0.130) in column (1); 0.831*** (0.133) in column (2).
  - Post Jul 2018 × Upstream Tariff: 1.160*** (0.393) in column (2); -0.349 (0.419) in column (3).
  - Post Jul 2018 × Downstream Tariff: 0.998*** (0.182) in column (3); 0.664*** (0.170) in column (4).
  - Lagged ln(FDI): ~0.822*** (0.015) across specifications.
- Table 4 (manufacturing vs non-manufacturing, Dependent variable: ln(FDI), N. of observations 45,678, R-squared 0.953):
  - Manufacturing × Post Jul 2018 × Output Tariff: 0.269** (0.109) in column (1); 0.485** (0.188) in column (2).
  - Manufacturing × Post Jul 2018 × Upstream Tariff: 0.931** (0.447) in column (1); -0.476 (0.492) in column (2).
  - Manufacturing × Post Jul 2018 × Downstream Tariff: 0.951*** (0.254) in column (1); 1.195*** (0.393) in column (2).
  - Non-manufacturing × Post Jul 2018 × Downstream Tariff: 0.619** (0.287) in column (1); 0.570** (0.242) in column (2).
  - Manufacturing lagged ln(FDI): 0.877*** to 0.879*** (0.017). Non-manufacturing lagged ln(FDI): 0.759*** to 0.757*** (0.021-0.022).
- Interpretation of results:
  - FDI to Mexico increased more in industries affected by the 2018-19 U.S. tariffs on Chinese products.
  - Statistically significant downstream tariff effects suggest indirect spillovers via input-output linkages, including to non-tradable sectors.
  - Upstream tariffs show mixed/insignificant impacts in several specifications.
  - Since 2017, FDI to northern states increased more than to southern states, aligning with concentration of firms exporting to the U.S.

### Conclusions and policy implications
- Mexico can capitalize on deeper integration with the North American market and ongoing reshaping of global value chains, given:
  - A developed manufacturing system.
  - Availability of human capital.
  - Cost competitiveness.
  - Proximity to the U.S. market.
- FDI has been instrumental in expanding Mexico’s productive capacity to supply U.S. demand; much of that FDI originates from the U.S.
- To fully exploit these favorable dynamics amid uncertainty, recommended priorities (as outlined in the 2024 Mexico Article IV Consultation Staff Report) include:
  - Improving environment for private investment and business development.
  - Opening markets to private sector participation.
  - Introducing regulatory reforms.
  - Broadening access to financial services.
  - Better-targeted public investment to relieve infrastructure bottlenecks, notably in energy, transport, and water.
  - Maintaining open trade policies to underpin business investment and bolster international competitiveness.

*Source: https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025180-source-pdf.pdf*

### Annex I.

### Annex I.

### Summary statistics: U.S. and Mexico imports (Table A.1; Table A.2)
- U.S. aggregate annual imports (Table A.1):
  - All: Annual Sum 2,220.3 (Mean 0.037, Obs. 59,970) and 2,872.4 (Mean 0.046, Obs. 62,208).
  - All Excluding China: Annual Sum 1,694.3 (Mean 0.028, Obs. 59,970) and 2,424.4 (Mean 0.039, Obs. 62,208).
  - China: Annual Sum 526.0 (Mean 0.011, Obs. 46,019) and 448.0 (Mean 0.010, Obs. 46,026).
  - Mexico: Annual Sum 317.2 (Mean 0.010, Obs. 32,723) and 480.1 (Mean 0.014, Obs. 35,467).
- Selected country-level annual sums and means (Table A.1, 2017 and 2023 columns apparent):
  - Canada: 305.9 (Mean 0.008, Obs. 40,402) and 429.6 (Mean 0.010, Obs. 42,660).
  - Japan: 139.8 (Mean 0.004, Obs. 33,632) and 151.6 (Mean 0.005, Obs. 33,129).
  - Germany: 120.0 (Mean 0.003, Obs. 37,628) and 163.0 (Mean 0.004, Obs. 38,001).
  - Ireland: 49.0 (Mean 0.006, Obs. 8,230) and 82.7 (Mean 0.009, Obs. 8,916).
  - Vietnam: 48.4 (Mean 0.003, Obs. 15,429) and 118.9 (Mean 0.006, Obs. 20,920).
  - South Korea: 73.4 (Mean 0.003, Obs. 26,251) and 119.7 (Mean 0.004, Obs. 27,402).
  - Mexico and other Latin American economies (selected): Brasil 30.6 (Mean 0.002, Obs. 16,526) and 41.0 (Mean 0.002, Obs. 18,681); Chile 11.3 (Mean 0.002, Obs. 4,722) and 16.6 (Mean 0.003, Obs. 4,944).
- Mexico-specific summary statistics (Table A.2):
  - All: Annual Sum 399.2 (Mean 0.010, Obs. 39,858) and 564.2 (Mean 0.0140, Obs. 40,187).
  - United States: Annual Sum 194.9 (Mean 0.006, Obs. 34,871) and 255.9 (Mean 0.0075, Obs. 34,087).
  - China: Annual Sum 74.1 (Mean 0.003, Obs. 26,478) and 114.2 (Mean 0.0039, Obs. 29,196).
  - Selected partners (Mexico imports/exports): Germany 16.4 (Mean 0.001, Obs. 15,788) and 21.2 (Mean 0.0014, Obs. 14,725); Vietnam 4.6 (Mean 0.001, Obs. 4,228) and 11.6 (Mean 0.0018, Obs. 6,440); Brazil 5.4 (Mean 0.001, Obs. 5,427) and 13.4 (Mean 0.0024, Obs. 5,587).
- Data sources and notes:
  - Sources: UNComtrade, Haver Analytics, INEGI, Banxico, National Authorities, and IMF staff calculations (as cited in tables and figures).
  - Note: UNComtrade data at montly and HS6-digit code level. Sample period referenced in regressions: 2016M1-2023M12.

### Estimated impact of 2018–19 U.S. tariffs on exports to the U.S. (Table A.3; pooled and augmented results Tables A.4–A.5)
- Econometric specification (Table A.3 notes):
  - ln(US imports)it = β Δtariffi × 1(post Jul 2018)t + μi + λt + εit.
  - Product and time fixed effects included; SEs clustered by HS6-digit code.
  - Sample period: 2016M1-2023M12.
- Country-level coefficients: Tariff increase × Post Jul 2018 (Table A.3)
  - MEX: 0.884*** (SE 0.222)
  - CAN: 0.168 (SE 0.148)
  - GTM: 1.534*** (SE 0.470)
  - HND: 1.486** (SE 0.623)
  - IDN: 1.089*** (SE 0.328)
  - MYS: 1.682*** (SE 0.424)
  - VNM: 0.662** (SE 0.336)
  - THA: 1.126*** (SE 0.259)
  - AUT: 0.655*** (SE 0.242)
  - POL: 1.196*** (SE 0.304)
  - Number of observations by country (selected): MEX 257,951; CAN 316,410; VNM 133,143; THA 142,833.
  - R-squared (selected): MEX 0.888; CAN 0.894; VNM 0.813.
  - Significance: *** p<0.01, ** p<0.05, * p<0.1.
- Pooled-sample regression with lag (Table A.4 notes):
  - ln(US imports)ict = β Δtariffi × 1(post Jul 2018)t + ρ ln(US imports)ict−1 + μic + λt + εit.
  - Lagged ln(US imports) coefficient in pooled sample: 0.446*** (SE 0.004) and 0.442*** (SE 0.004) in reported variants.
  - N. of observations pooled: 1,172,707; R-squared 0.885.
- Impact on Mexico’s imports by type of goods (Table A.5; dependent variable ln(MEX imports)):
  - Common coefficient: Tariff increase × Post Jul 2018 = 0.311*** (SE 0.068).
  - Lagged ln(MEX imports) = 0.363*** (SE 0.009).
  - N. of observations: 183,451 (per-country panels vary; see table).
  - By goods type and country: Tariff increase × Post Jul 2018 × Intermediate goods (selected coefficients):
    - CHN: 0.335*** (SE 0.113)
    - USA: 0.185* (SE 0.095)
    - BRA: -0.343 (SE 0.425)
    - CHL: 2.258* (SE 1.155)
    - COL: 2.008*** (SE 0.745)
    - IDN: -0.202 (SE 1.794)
    - MYS: 0.708 (SE 0.526)
    - VNM: 0.365 (SE 0.512)
    - THA: 0.903 (SE 0.843)
    - JPN: 0.529 (SE 0.521)
    - KOR: 0.276 (SE 0.415)
    - DEU: 0.448** (SE 0.190)
    - ESP: 0.459 (SE 0.316)
    - POL: 0.369** (SE 0.177)
    - PRT: 0.557** (SE 0.238)
    - PR T (likely Portugal) and other entries: 2.999*** (SE 0.676) appears for the last column in that panel.
  - By goods type: Tariff increase × Post Jul 2018 × Consumption goods and × Capital goods coefficients reported (selected examples):
    - CHN Consumption goods: 0.316*** (SE 0.114); Capital goods: 0.491*** (SE 0.105).
    - USA Consumption goods: -0.085 (SE 0.113); Capital goods: -0.067 (SE 0.104).
    - VNM Consumption goods: 0.143 (SE 0.658); Capital goods: 0.123 (SE 0.460).
  - Lagged ln(MEX imports) by country (selected):
    - CHN: 0.363*** (SE 0.009)
    - USA: 0.362*** (SE 0.010)
    - BRA: 0.293*** (SE 0.017)
    - VNM: 0.352*** (SE 0.021)
    - JPN: 0.419*** (SE 0.021)
    - DEU: 0.378*** (SE 0.016)
    - POL: 0.229*** (SE 0.012)
  - Sample period: 2016M1-2023M12. Standard errors clustered by HS6-digit code. Country-by-product and time fixed effects included. *** p<0.01, ** p<0.05, * p<0.1.

### Figures: trade shares, industry composition, and dynamics (Figures A.1–A.14)
- U.S. exports and imports by country (Figure A.1; percent shares over 2007–2023):
  - Shifts shown for Mexico, Canada, China, UK, Euro Area, India, Germany, Japan, South Korea, and Rest of the world across 2007–2023.
- Mexico: imports and exports by country (Figure A.2; percent shares over 2007–2023):
  - Panels show Mexico, Spain, Canada, Germany, China, Rest of the World and breakdowns over 2007–2023 for exports and imports.
- Mexico and China: exports and imports by industry (Figures A.3–A.4; percent of total by industry, 2007–2023 and 2000–2023 for China panels):
  - Industry categories used: Agriculture, Oil, Transportation, Machinery, Electronics, Electrical, Miscellaneous, Chemicals, Food, Other; and for China also Manufacturing Commodities.
- Mexico manufacturing exports and vehicle export contributions (Figures A.5–A.6):
  - Index (2017=100) for Total Manufacturing Exports (100%) and subcomponents: Automotive (36%), Machinery and Equipment (15%), Electronic (17%), Professional and Scientific Equip. (5%).
  - Contributions to vehicle export growth shown by partner: United States, Canada, Germany, Brazil, Colombia, Others for 2017–2023 (In percent, Y/Y).
- Mexico: imports from China and exports to the U.S. by goods type (Figure A.7; In USD millions) — percentile change ranges shown:
  - FDI percentiles (legend ranges): 0.6 - 6.1; 0.2 - 0.6; 0.1 - 0.2; -0.1 - 0.1; -0.1 - -0.1; -0.3 - -0.1; -0.4 - -0.3; -0.9 - -0.4; -1.8 - -0.9.
  - FDI and Exports percentiles (legend ranges): 17.9 - 26.5; 8.8 - 17.9; 4.9 - 8.8; 4.3 - 4.9; 2.1 - 4.3; 1.1 - 2.1; 0.7 - 1.1; 0.2 - 0.7; -0.3 - 0.2.
- Mexico: FDI and exports 2017–2023 (Figure A.8; In USD billions):
  - FDI and Exports panels plotted for 2017–2023; percentiles reflect relative change in FDI and Exports between 2017 and 2023.
- FDI to Mexico by sector and by country/industry (Figures A.9–A.10; In USD billions and percent):
  - FDI by industry (selected): Manufacturing, Finance, Accommodation, Mining, Retail trade, Transportation, Professional serv., Real estate, Construction, Agriculture for 2022 and 2023.
  - Foreign Direct Investment by Industry (percent) and by Country (percent) time-series 2007–2023: US, Spain, Canada, Germany, China, Rest of the World.
  - Note: “Other industries include Scientific, Agriculture, Cultural, Other Social Services, Health, Business and Education.”
- Mexico exports by region (Figure A.11; In USD billions):
  - Regional split: South, Center, North (North states include AG, BC, BCS, COAH, CL, CI, DUR, JAL, MIC, NAY, NL, SLP, SIN, SON, TAM and ZAC).
- HS6-digit level analysis: dynamics and treatment effects (Figures A.12–A.14):
  - Dynamic log-difference plots (2016–2023) show treatment vs control paths for US imports from IDN, MYS, VNM, THA, POL, and All.
  - Key event markers shown on time series: Tariff Phase 1–4, COVID-19, Russia's invasion of Ukraine.
  - Treatment definition: tariff increase >=15% (Treatment) vs tariff increase <15% (Control).
  - Treatment effect by periods (Figure A.13): Post Tariff, Post COVID-19, Post War in Ukraine panels for countries CHN, MEX, GTM, HND, IDN, MYS, VNM, THA, AUT, POL with scales differing by country (examples: CHN axis -2 to 0.8; MEX axis 0 to 2.5; GTM axis -1 to 3).
  - Treatment effect by categories (Figure A.14): category-level treatment effects plotted for China, Mexico, and pooled/all excluding China across categories:
    - Categories listed: Animal & Animal Products; Vegetable Products; Foodstuffs; Mineral Products; Chemicals & Allied Industries; Plastics / Rubbers; Raw Hides, Skins, Leather, & Furs; Wood & Wood Products; Textiles; Footwear / Headgear; Stone / Glass; Metals; Machinery / Electrical; Transportation; Miscellaneous.
    - Scales: China panel spans -2.5 to 0; Mexico panel spans -1 to 1; All excluding China panel spans -1 to 1.

### Key methodological notes and data specifics
- Regression specifications:
  - Table A.3: product and time fixed effects; ln(US imports) explained by Δtariff × post indicator.
  - Table A.4: pooled sample adds lagged ln(US imports) and product-by-country fixed effects.
  - Table A.5: regression augmented to allow different coefficients across goods type; raw materials excluded due to small sample.
- Standard errors:
  - SEs clustered by HS6-digit code across regressions; significance levels reported as *** p<0.01, ** p<0.05, * p<0.1.
- Sample period for regression analyses: 2016M1-2023M12.
- Data sources cited within tables/figures: UNComtrade, Haver Analytics, INEGI, Banxico, National Authorities, and IMF staff calculations.

*Source: Annex I (Tables A.1–A.5; Figures A.1–A.15) from the IMF Working Paper chapter: "Relocation of Global Value Chains: The Role of Mexico" (Annex I content provided).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2025/english/wpiea2025180-source-pdf.pdf_