## CHAPTER 4

## Source details

**Canonical URL:** [CHAPTER 4](https://www.imf.org/-/media/files/publications/weo/2019/april/english/ch4.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/weo/2019/april/english/ch4.pdf.md)
- [Structured JSON version](/-/media/files/publications/weo/2019/april/english/ch4.pdf.json)

---

### Introduction and context
- "In both advanced and emerging market economies, 'more than 80 percent of the public views trade in a positive light—yet, fewer than half of these people are convinced that trade benefits jobs, wages, or prices.'"
- Trade delivers gains by enabling specialization according to comparative advantage, enhancing competition, and enabling knowledge and technology flows, while widening choice and lowering prices.
- Trade imposes adjustment costs on some workers and communities; "open and fair trade, with lower or no tariffs or other obstacles to trade, can bring lasting net benefits to all involved if the right policies are in place to ensure that the gains are widely shared and those bearing the brunt of adjustment receive the help they deserve."

### Research questions and analytical approach
- Research questions:
  - What drives bilateral trade balances—macroeconomic factors versus tariffs and micro-structural determinants affecting comparative advantage and the international division of labor?
  - What links aggregate trade balances (and their drivers) to bilateral balances?
  - What are the consequences and spillovers when bilateral tariffs are raised, accounting for global value chains (GVCs)?
- Analytical approach:
  - Uses the gravity model to explain bilateral exports as a function of trade costs (including tariffs), aggregate supply and demand, and sectoral composition.
  - Focuses on changes over time (average 2010–15 minus average 1995–99) to exploit variation not captured by time-invariant gravity determinants.
  - Data: OECD Trade in Value Added database, 63 countries, ISIC 3, 34 sectors, 1995–2015; nominal US dollars; estimation over five-year periods.
  - Limitations: results interpreted as partial-equilibrium effects; macroeconomic factors capture many aggregate drivers (demographics, institutions, macro policies, supply-side policies) that are hard to separate.

### Key quantitative findings
- Trade-cost elasticity:
  - A 1 percent increase in gross ad valorem tariffs reduces gross bilateral exports by about 3–6 percent.
- Decomposition (1995–2015):
  - Macroeconomic factors are by far the largest contributors to changes in bilateral trade balances; trade costs contribute to a lesser degree.
  - Example: If Chinese tariffs on US goods had remained at their 1995 level, the US–China trade deficit would have been, on average, $30 billion (about 12 percent) larger.
- Scaling and bilateral amplification:
  - Under mild assumptions, bilateral trade balance depends on relative aggregate trade-balance-to-GDP ratios, sizes relative to the world economy, and bilateral trade intensities.
  - Example magnitude: A 0.5 percentage point exogenous decrease in Germany’s overall trade balance—holding everything else constant—would reduce:
    - Germany’s bilateral trade balances with China, the Netherlands, and the United States by about 0.1 percent of Germany’s GDP,
    - Germany’s bilateral trade balances with the United Kingdom and France by about 0.05 percent of Germany’s GDP.

### Role of macroeconomic factors (EBA determinants and findings)
- EBA determinant groups:
  - Macroeconomic policies: fiscal policy (cyclically adjusted fiscal balance) and exchange rate policy (foreign exchange interventions).
  - Credit: detrended private credit to GDP.
  - Cyclical: temporary cyclical factors such as the output gap and commodity terms of trade.
  - Fundamentals: demographics, level of economic and institutional development, social safety nets, commodity reliance, net foreign asset position, role as safe/reserve asset provider.
- Empirical findings for 2010–17 (examples):
  - Credit: contributed to reduction in external imbalances (United States: positive contribution after correction of pre-crisis credit boom; China: post-crisis credit expansion increased domestic demand and reduced external surplus).
  - Fiscal policy:
    - Germany: tight fiscal policy contributed to a large trade surplus.
    - Japan: relatively loose fiscal stance contributed to balancing external trade position.
    - United States: expansionary fiscal stance after the global financial crisis offset other improvements and recent fiscal stimulus projected to further widen the trade deficit.
  - Foreign exchange interventions: limited impact on aggregate trade balances; largely disappeared for China in recent years.
  - Cyclical factors: provided, on average, a significant negative component in all countries except Japan.
  - Fundamentals: level of development made a positive contribution for advanced economies and a negative contribution for China; demographics make a negative contribution in late-stage aging countries.
  - Unexplained residual: quantitatively significant, likely reflecting macroeconomic distortions not captured (supply-side policies such as production subsidies and regulatory policies).

### Role of sectoral specialization
- Sectoral composition changes matter nontrivially for many bilateral imbalances.
- Positive sectoral effect on a bilateral trade balance implies either:
  - Output share of sectors where a country had large initial exports rose (relative to spending), or
  - Spending share of sectors where the country had high initial demand fell (relative to output).
- Empirical examples:
  - US sectoral changes contributed positively to bilateral balances with China, Germany, and Japan.
  - Germany’s sectoral changes contributed positively to bilateral balances with Italy and the United States.

### Role of trade costs and tariffs
- Declines in bilateral tariffs boosted gross trade but had muted effects on net trade balances because:
  - Tariffs were already low in the mid-1990s in many countries.
  - Tariff reductions were often reciprocal, producing offsetting effects.
- Tariffs shape the international division of labor and influence how macroeconomic factors map into bilateral trade.
- Tariffs have important long-term effects on productivity, output, and employment by enabling or limiting access to foreign inputs and competition.

### Measuring tariff spillovers in a GVC context (constructed measures)
- Four constructed measures:
  - Upstream tariff: average cumulative tariff applied to intermediate inputs as a share of sector output — captures effect on production costs.
  - Domestic protection tariff: average import-weighted tariff imposed on imports that compete with domestic sector output — affects domestic demand.
  - Downstream tariff: average cumulative tariff the sector’s output faces when exported (directly or indirectly) — affects international demand.
  - Diversion tariff: captures trade-diversion benefits when competitors face tariffs.
- Construction uses global input-output matrices and follows theoretical and empirical methods cited in the chapter.

### Empirical findings on tariffs and value chains (manufacturing panel)
- Sample: 35 countries, 13 manufacturing sectors, controlling for country-specific macro changes and country-industry characteristics.
- Main conclusions:
  - Upstream and downstream tariffs matter more for output and productivity than domestic protection tariffs.
  - Upstream and downstream tariffs have statistically significant negative effects on real value added, employment, labor productivity, and total factor productivity.
    - Mechanisms: upstream tariffs raise input costs; downstream tariffs reduce international demand; both lower returns to scale from international market participation.
  - Domestic protection tariffs do not show significant positive productivity effects; small negative effect on employment observed.
  - Evidence of trade diversion: diversion tariff is positive and statistically significant for value added and employment (firms benefit when competitors face tariffs); diversion tariff is insignificant for labor productivity and total factor productivity.
- Summary sign patterns (from Table 4.1):
  - Real Value Added: Upstream Tariff (negative), Domestic Protection (positive), Downstream Tariff (negative), Diversion Tariff (positive).
  - Number Employees: Upstream Tariff (negative), Domestic Protection (negative), Downstream Tariff (negative), Diversion Tariff (positive).
  - Labor Productivity: Upstream Tariff (negative), Domestic Protection (insignificant), Downstream Tariff (negative), Diversion Tariff (insignificant).
  - Total Factor Productivity: Upstream Tariff (negative), Domestic Protection (positive), Downstream Tariff (negative), Diversion Tariff (insignificant).

### Simulations and amplification by global value chains
- Simulations are partial-equilibrium illustrations using empirical coefficients; do not capture confidence effects, policy uncertainty, or financial channels.
- 2011 chosen as simulation base year (closest recent GVC links available).
- Greater GVC integration amplifies sensitivity of upstream and downstream tariffs to nominal tariff changes; a 1 percentage point tariff increase by all countries has a larger negative effect in 2011 than in 1995.
- Countries with large, highly integrated manufacturing sectors (Germany, Korea, China, Japan) experience larger negative GDP effects from nominal tariff hikes than countries with smaller, less integrated manufacturing sectors (Canada, United States).

### Bilateral discriminatory tariff scenarios and third-country effects
- Discriminatory 1 percentage point reciprocal tariff increase between China and the United States:
  - China and the United States are both hurt:
    - China: downstream and, to a slightly lesser extent, upstream tariffs dominate.
    - United States: upstream tariffs are more important due to intermediate imports from China.
  - Third-country effects:
    - Trade diversion offsets negative spillovers for many third countries.
    - Mexico and Canada benefit most from substitution of China’s exports to the United States.
    - East Asia benefits to some extent; Japan and Korea face downstream-affected channels but also benefit from diversion.
    - Canada likely sees an overall positive effect due to trade diversion.
  - The percent-of-GDP cost is smaller for the United States because manufacturing is a smaller share of US GDP; manufacturing value-added change is somewhat larger in absolute terms for the United States than for China.
- General equilibrium large US–China dispute (three models summarized below):
  - China and the United States suffer largest losses; collapse in bilateral trade and lower returns to capital.
  - Trade diversion toward Mexico, Canada, and east Asia occurs.
  - No economically significant change in each country’s aggregate trade balance despite large bilateral reductions.
  - Macroeconomic spillovers in third countries are modest; sectoral spillovers (repositioning of GVCs) are larger.
  - Long-term sectoral shifts: sizable manufacturing relocation away from China and the United States toward Mexico, Canada, and east Asia, with sizable sectoral job losses in China and the United States.
  - Broader effects: increased uncertainty, negative confidence effects, and tightened global financial conditions would make most countries worse off even if some benefit from trade diversion.

### Macro modeling of a hypothetical 25 percentage point US–China tariff increase (Box 4.4)
- Scenario: tariffs on all US–China goods trade increase by 25 percentage points.
- Models used: GIMF, GTAP, CFRT (heterogeneous-firm model with entry and exit).
- Trade collapse:
  - US–China trade falls by 25–30 percent in the short term (GIMF).
  - US–China trade falls between 30 percent and 70 percent over the long term (model-dependent).
- Annual real GDP losses:
  - United States: –0.3 percent to –0.6 percent.
  - China: –0.5 percent to –1.5 percent.
- Asymmetry: China typically incurs larger effects because exports to the United States represent a larger share of China’s economy.
- Multilateral trade balance: no economically significant change in each country’s multilateral trade balance (GIMF result).
- Third-country effects:
  - Third countries increase exports to the United States and decrease exports to China; Mexico and Canada benefit most.
  - Model differences: GIMF shows largest short-term increase in third-country exports; CFRT shows sizable positive effects for some countries; GTAP shows weaker/sometimes negative responses.
- Sectoral reallocations and labor markets:
  - Large worldwide manufacturing contraction; manufacturing expands in Mexico, Canada, and some Asian countries.
  - Services expand in China, contract in several other countries.
  - US agricultural sector: sizable contraction.
  - Job loss examples (CFRT):
    - United States: about 1 percent of the workforce lost in agricultural and transportation equipment sectors.
    - China: about 5 percent of the workforce lost in other manufacturing sector.
- Repositioning of GVCs (electronics and machinery example):
  - Electronics and machinery are about 20 percent of world imports.
  - CFRT results: China loses position as top exporter of electronics and machinery to the United States; East Asia, Mexico, and Canada gain market shares (examples of percentage shifts in export shares reported in the box).

### Gross versus value-added trade balances
- Gross bilateral imbalances often overstate value origins in GVC contexts; value-added measures are typically smaller than gross measures.
- Large gross imbalances are generally accompanied by large value-added imbalances, but value-added imbalances shrink relative to gross measures because intermediates embody value from multiple countries.
- Panel evidence for 2015: the 10 largest gross bilateral imbalances are generally smaller in value-added terms; US deficits with Mexico and China shrink in value-added terms; aggregate trade balance of a country is identical whether using gross or value-added measures when measured against the totality of the rest of the world.

### Conclusions and policy recommendations
- Core conclusions:
  - Aggregate imbalances tend to be reflected at the bilateral level; bilateral imbalances are of little consequence for the aggregate: targeting particular bilateral trade balances without addressing macroeconomic conditions is likely to lead to trade diversion and offsetting adjustments with other partners.
  - Over the past two decades, macroeconomic factors (fundamentals, macro policies, credit cycles, and sometimes exchange rate and supply-side policies) were key drivers of bilateral balance evolution.
  - Declines in tariffs and trade costs enabled deeper GVCs, specialization by comparative advantage, and more efficient cross-border production.
  - Higher tariffs would significantly harm value added, employment, and productivity in affected countries and third countries through value-chain links; greater international division of labor increases the scope for negative spillovers.
  - Some countries may gain from trade diversion, but escalation of trade tensions raises uncertainty, lowers confidence, and tightens global financial conditions, making most countries worse off.
- Two main policy recommendations:
  - Focus on aggregate trade balances and current accounts: policy discussion should emphasize macroeconomic imbalances and the macroeconomic policies and distortions that drive them. Aggregate external imbalances are not inherently bad but distortive policies that create excessive and unsustainable imbalances should be avoided.
  - Lower tariffs broadly: multilateral reductions in tariffs and nontariff barriers boost trade and enable the international division of labor to better reflect comparative advantage, yielding output, employment, and productivity gains up and down value chains.
- Distributional policies:
  - Implement policies to ensure gains from trade are widely shared and to protect those harmed by structural adjustment, including:
    - Enhancing social safety nets (modern income support programs and unemployment assistance).
    - Retraining and reintegration programs.
    - Tax and benefit policy changes to redistribute gains from trade more evenly.

*Source: CHAPTER 4 — ThE DRIvERS OF BILaTERaL TRaDE aND ThE SPILLOvERS FROM TaRIFFS, International Monetary Fund | April 2019*

### conclusions. First, the discussion of external imbalances (of

### ch4 - conclusions. First, the discussion of external imbalances (of

### Introduction and context
- In both advanced and emerging market economies, "more than 80 percent of the public views trade in a positive light—yet, fewer than half of these people are convinced that trade benefits jobs, wages, or prices."
- Skepticism about trade is particularly pronounced in advanced economies.
- Trade provides gains by enabling specialization according to comparative advantage, enhancing competition, and enabling knowledge and technology flows; it also benefits consumers by widening choice and lowering prices.
- Trade can impose adjustment costs on some workers and communities; the chapter stresses that "open and fair trade, with lower or no tariffs or other obstacles to trade, can bring lasting net benefits to all involved if the right policies are in place to ensure that the gains are widely shared and those bearing the brunt of adjustment receive the help they deserve."

### Research questions addressed
- What drives bilateral trade balances—specifically, the role of macroeconomic factors compared with tariffs and micro-structural determinants that affect comparative advantage and the international division of labor?
- What is the link between aggregate trade balances (and their drivers) and bilateral balances?
- What are the consequences when bilateral tariffs are raised, and what spillovers arise for others when accounting for global value chains?

### Analytical approach
- Uses the gravity model for bilateral trade flows to explain changes in bilateral trade balances over time, disentangling the impact of trade costs (including tariffs), the international division of labor, and macroeconomic factors.
- Focuses on changes over time because gravity-model variables do not capture all time-invariant determinants of bilateral levels.
- Macroeconomic factors include aggregate supply and demand determinants: macroeconomic policies, fundamentals such as demographics, and longer-term effects of tariff changes and supply-side policies.

### Key findings
- Overall trade balances matter more than bilateral trade balances.
  - "Changes in overall (that is, aggregate) trade balances tend to affect most bilateral trade balances while—absent changes in macroeconomic conditions—a change in a bilateral trade balance tends to be offset by changes in bilateral balances with other trading partners, with little or no impact on the overall trade balance."
- Evolution of bilateral trade balances over the past two decades was, to a significant extent, driven by macroeconomic factors:
  - Drivers included demographics and the level of economic and institutional development; macroeconomic policies, in particular fiscal policy and credit cycles; and in some cases exchange rate policies and domestic supply-side policies (for example, subsidies to production costs).
- Changes in bilateral tariffs played a smaller role in the evolution of bilateral trade balances:
  - Reflects already-low starting tariff levels in many countries and simultaneous tariff reductions across countries, creating offsetting effects on net trade.
  - Nonetheless, tariff levels are an important part of bilateral trade costs shaping the international division of labor and affecting how macroeconomic factors influence bilateral trade.
- Tariffs have important effects on productivity, output, and employment over the longer term:
  - "The decline of tariffs to lower levels enabled a greater international division of labor—including through global value chains—and enhanced competition and access to foreign inputs, resulting in strong productivity improvements."
  - Increases in bilateral tariffs will hurt output, employment, and productivity in affected economies and in bystanders up and down value chains; while some countries may benefit from trade diversion, higher tariffs would leave the global economy worse off.

### Policy conclusions and recommendations
- Two main policy conclusions:
  - Discussion of external balances should focus on macroeconomic determinants of trade and current account balances. Changes in macroeconomic policies (for example, fiscal policy) will affect all bilateral balances. Absent addressing macroeconomic conditions, targeting particular bilateral trade balances will likely result only in trade diversion and offsetting changes with other partners.
  - Broad-based, multilateral reductions in tariffs and other nontariff barriers will benefit trade and, over the longer term, improve macroeconomic outcomes. Reductions in tariffs enable efficiency and dynamic gains by allowing greater specialization according to comparative advantage, integrating into supply chains, and improving access to foreign inputs.
- Recognize distributional concerns:
  - Put in place policies to ensure the gains from trade are widely shared and that those left behind are adequately protected.
  - Policies to help those harmed by structural adjustment include enhancing social safety nets (modern income support programs and unemployment assistance), retraining and reintegration programs, and tax and benefit policy changes to redistribute gains from trade more evenly.

### Stylized facts on trade patterns and global integration
- Aggregate and bilateral balances are highly related: countries with large overall trade imbalances tend to have more one-sided bilateral balances (examples cited include Germany and the United States).
- There is wide variation in bilateral trade balances across partners; countries with small overall balances can have large, offsetting bilateral balances.
- Global bilateral and overall trade balances evolved broadly in parallel over the past two decades, increasing strongly up to the 2008–09 global financial crisis and dipping during the crisis.
- Global trade integration was fostered by a persistent fall in trade costs:
  - Between 1995–99 and 2010–15, trade became relatively easier for most country pairs.
  - Improvements in transportation technologies and policy changes (for example, expansion of World Trade Organization membership and generalized tariff declines) reduced barriers.
  - Reductions were particularly marked where tariffs were initially high, as in China and other emerging market economies.
- Comparative advantage and specialization evolved alongside declining trade costs, reinforcing manufacturing specialization in countries with revealed comparative advantage (China, Korea, Germany, Japan, and Mexico) and relative de-specialization in manufacturing in countries with initial comparative disadvantage (United States, United Kingdom).
- Global value chains deepened specialization and amplified multicountry trade patterns:
  - Production organized around three poles: North America (centered on the United States), Europe (centered on Germany), and Asia (with China playing a central role by 2015).
  - Greater participation in global value chains is strongly positively related to the size of a country's bilateral balances but much less so to the size of the overall trade balance.
- The difference between gross trade measures and value-added measures increased as goods incorporate value added from multiple countries along supply chains.

*Source: ch4 - conclusions. First, the discussion of external imbalances (of*

### CHAPTER 4

### CHAPTER 4

### The Gravity Model and Methodology
- The chapter uses the gravity model as the workhorse model of the trade literature to explain bilateral exports as a function of trade costs, aggregate supply and demand of trade partners, and the sectoral composition of demand and supply.
- The gravity model explains bilateral exports through three sets of determinants:
  - Macroeconomic factors: bilateral exports increase with the aggregate supply (gross output) of the exporting country and the aggregate demand (gross spending) of the importing country, scaled by world output. The analysis uses gross output and spending (instead of value added and final spending) to account for growth in global-value-chain-related intermediates’ trade, which is included in export measures.
  - Trade costs: include natural trade costs and man-made—or policy-related—trade costs. The model controls for bilateral tariffs and average (multilateral) trade costs faced by the exporter and importer to capture that the effect of bilateral tariffs is relative to trade costs with other partners.
  - Sectoral composition of supply and demand: sectoral specialization affects how much two countries trade in various sectors and hence in the aggregate. The model is estimated at the sectoral level and aggregated to the country level to identify these effects.
- Data and estimation:
  - Uses the Organisation for Economic Co-operation and Development’s Trade in Value Added database.
  - Database reports bilateral export data for 63 countries at the International Standard Industrial Classification of All Economic Activities 3 level for 34 sectors from 1995 to 2015.
  - All variables are expressed in nominal US dollars.
  - Estimation is carried out over five-year periods to remove short-term effects of nominal variables, such as nominal exchange rate movements.
- Limitations noted:
  - Results are best interpreted as partial-equilibrium effects and not necessarily as the complete dynamic interaction of trade, macroeconomic factors, and tariffs over time.
  - Macroeconomic factors capture all factors and policies that impact aggregate supply and demand, including demographics, institutional development, macroeconomic policies, and supply-side policies; these are difficult to identify separately given lack of consistent measures across countries.
  - The gravity model loses explanatory power when applied to the level of bilateral trade balances; the chapter therefore focuses on explaining changes in bilateral trade balances.

### Key Quantitative Findings from the Gravity Estimates
- Trade-cost elasticity:
  - The estimated elasticity implies that a 1 percent increase in gross ad valorem tariffs reduces gross bilateral exports by about 3–6 percent.
- Data coverage and specification robustness:
  - Country-level results are highly robust to the sector-level specification.
  - Important differences exist between services and non-services sectors; distance is a significant hindrance to both types of trade but is more important for trade in services.
- Model fit:
  - The model explains bilateral exports (and imports) very well across all specifications.

### Decomposing Changes in Bilateral Trade Balances (1995–2015)
- Approach:
  - Contributions of trade costs, macroeconomic factors, and sectoral composition to changes in bilateral trade balances are computed using the estimated model and presented for major trading partners of the United States, China, and Germany.
  - The decomposition examines average value 2010–15 minus average value 1995–99.
- Overall results:
  - Macroeconomic factors are, by far, the largest contributors to changes in bilateral trade balances over the period 1995–2015.
  - Trade costs contribute as well, though to a lesser degree.
  - The precise impact of macroeconomic factors on bilateral trade balances depends on the initial state of the bilateral relationship (for example, whether the bilateral balance was large and positive or negative).
- Example magnitude:
  - If Chinese tariffs on US goods had remained at their 1995 level, the estimation suggests that the US–China trade deficit would have been, on average, $30 billion (about 12 percent) larger.

### Role of Macroeconomic Factors
- Domestic macroeconomic conditions:
  - Reflect the evolution of gross output and gross spending in a country.
  - The magnitude of impact depends on the initial bilateral trade balance, trade costs, international division of labor, initial macroeconomic conditions, and other structural factors.
  - Over 1995–2015:
    - US domestic macroeconomic factors had a negative effect on its bilateral trade balances across trading partners because US gross output was growing more slowly than spending (the United States was, in the aggregate, spending more than it was producing).
    - Germany’s domestic macroeconomic factors had a positive effect on its bilateral trade balances, reflecting faster growth of output than spending.
    - China’s output was growing faster than spending, in part reflecting domestic supply-side policies, such as subsidies to the cost of production of manufactured (traded) goods; where China initially had a bilateral surplus (for example, with the United States), faster growth of output than spending amplified the bilateral surplus.
- Foreign macroeconomic conditions:
  - The evolution of spending and output in partner countries matters; the initial structure of trade influences the resulting change in bilateral trade balances.
  - For Germany, foreign macroeconomic factors contributed to its bilateral surpluses, reflecting faster growth of spending relative to output in partner countries and Germany’s initial surplus with partners.

### Role of Sectoral Specialization
- Changes in the sectoral composition of aggregate demand and supply play a nontrivial role for many bilateral imbalances.
- A positive effect on a bilateral trade balance indicates:
  - The output share of sectors where a country had large initial exports rose (more than the spending share), supporting path dependence in production structures and comparative advantage; or
  - The spending share of sectors where the country had high initial demand fell (more than the output share).
- Empirical examples:
  - Sectoral changes in US supply and demand contributed positively to its bilateral balances with China, Germany, and Japan.
  - Germany’s sectoral changes contributed positively to its bilateral trade balances with Italy and the United States.

### Role of Trade Costs and Tariffs
- Declines in bilateral tariffs contributed to growth in the level of gross trade (exports and imports) but had a more muted impact on trade balances (net trade).
- Reasons for muted net effects:
  - Tariffs were already low in the mid-1990s in many countries.
  - Tariff reductions were often reciprocal, producing offsetting effects on bilateral trade balances.
- Example:
  - Changes in bilateral tariffs contributed slightly positively to changes in US bilateral trade balances with Canada, China, and Mexico because these countries’ tariffs on US goods were falling faster than tariffs imposed by the United States on their goods (albeit from a higher level), mechanically promoting a greater rise in US exports to these countries than in their exports to the United States.

*International Monetary Fund | April 2019 — CHAPTER 4*

### CHAPTER 4

### CHAPTER 4

### The Role of Macroeconomic Factors
- Changes in a country’s bilateral trade balances are driven significantly by changes in the imbalance between gross production and spending of each trading partner (the country’s aggregate trade balance).
- Under mild assumptions, any bilateral trade balance depends on:
  - the relative size of the two countries’ aggregate trade balance-to-GDP ratios,
  - the two countries’ sizes relative to the world economy,
  - bilateral trade intensities.
- Shrinking of the aggregate trade balances of the United States and China after the global financial crisis was matched by a corresponding contraction in their bilateral trade balance once aggregate and bilateral trade balances are scaled in a theory-consistent way.
- The way a change in a country’s aggregate trade balance is reflected in bilateral balances depends on partner-specific trade intensities. Example:
  - A 0.5 percentage point exogenous decrease in Germany’s overall trade balance—holding everything else constant—would reduce:
    - Germany’s bilateral trade balances with China, the Netherlands, and the United States by about 0.1 percent of Germany’s GDP,
    - Germany’s bilateral trade balances with the United Kingdom and France by about 0.05 percent of Germany’s GDP.
- Larger bilateral effects reflect either a higher trade intensity with Germany (e.g., the Netherlands) or a higher share in world output (e.g., the United States).

### EBA Determinants and Findings (Drivers of Aggregate Trade Balances)
- The IMF’s External Balance Assessment (EBA) framework relates the current account (and here, the trade balance) to macroeconomic policies and other drivers. Main EBA determinants grouped into four broad groups:
  - Macroeconomic policies: fiscal policy (cyclically adjusted fiscal balance) and exchange rate policy (foreign exchange interventions). Note: macroeconomic policy distortions (for example, wide-spread export or production subsidies) are difficult to measure and not explicitly captured.
  - Credit: captured by detrended private credit to GDP.
  - Cyclical: temporary cyclical factors such as the output gap and the commodity terms of trade.
  - Fundamentals: demographics, level of economic and institutional development, social safety nets, reliance on commodity exports, net foreign asset position, and role as provider of safe and reserve assets.
- Empirical findings for 2010–17 (examples: United States, Japan, Germany, China):
  - Financial conditions (credit) contributed to the reduction in external imbalances:
    - United States: contribution of credit conditions to the trade balance is now positive after correction of the credit boom before the global financial crisis.
    - China: credit expansion after the crisis increased domestic demand and reduced the external surplus.
  - Fiscal policy effects:
    - Germany: tight fiscal policy contributed to a large trade surplus.
    - Japan: relatively loose fiscal stance contributed to balancing the external trade position.
    - United States: relatively expansionary fiscal stance after the global financial crisis offset other improvements in the trade balance, and recent fiscal stimulus is projected to further widen the trade deficit.
  - Foreign exchange interventions had only a limited impact on aggregate trade balances and, for China, have largely disappeared in recent years.
  - Cyclical factors provided, on average, a significant negative component in all countries except Japan.
  - Among fundamentals:
    - Level of development made a positive contribution to the aggregate trade balance of advanced economies and a negative contribution for China.
    - Demographic variables make a negative contribution to the trade balance of countries in the late stage of the aging process.
  - The unexplained residual is quantitatively significant and likely reflects macroeconomic distortions not directly accounted for by EBA drivers, including supply-side policies (for example, production subsidies and regulatory policies that affect aggregate supply).

### Relationship between Bilateral and Overall Trade Balance Adjustments
- Empirical observation: large changes in overall trade balances tend to be accompanied by similar changes in a country’s bilateral trade relationships, while large adjustments in specific bilateral deficits do not necessarily lead to large adjustments in the overall trade balance.
- Absent changes in macroeconomic conditions, large changes in one bilateral trade balance of a country tend to result in compensating adjustments in other bilateral balances.

### A Closer Look at Tariffs and Their Spillovers
- Over 2010–15, changes in average trade costs faced by countries across all trading partners often played a larger role than changes in particular bilateral tariffs.
- Rising world output shares of emerging market economies, especially China, increased trade with those economies and reduced trade with some others; because those rapidly growing countries displayed higher tariffs than typical advanced economies, average trade costs rose and altered trade patterns—making trade between lower-cost country pairs more attractive and amplifying bilateral trade balances (for example, widening US bilateral deficits and increasing Germany’s bilateral surpluses).
- Compared with macroeconomic factors, changes in bilateral tariffs played a smaller role in the evolution of bilateral balances, but:
  - Large and persistent changes in tariffs can significantly impact the international division of labor, productivity, and macroeconomic factors in the longer term.
  - An increase in bilateral tariffs to prohibitive levels would cripple trade, whether bilateral or global, with severe consequences for the economies involved.
- Global value chain (GVC) context:
  - Reduction in tariffs and transportation and communications costs since the mid-1990s corresponded with a significant increase in complex global value chain participation (share of exports that crosses at least two borders).
  - Increased GVC participation creates scope for specialization and productivity improvement but increases the risk of international spillovers from tariff increases and trade wars.
  - Tariffs upstream (on suppliers) and downstream (on users/customers) can affect firms’ cost of production and demand for output; tariffs on competitors can affect relative competitiveness and demand (trade diversion effects).

### Measuring Spillovers from Tariffs (Constructed Measures)
- Four measures constructed to capture direct and indirect effects of tariffs on production, employment, and productivity in a global value chain context:
  - The upstream tariff:
    - Definition: the average cumulative tariff applied to the intermediate inputs as a share of the sector’s output.
    - Purpose: captures the average effect tariffs have on the cost of production.
    - Calculation: uses the global input-output matrix, with individual elements scaled by the relevant sectoral output.
  - The domestic protection tariff:
    - Definition: the average tariff (import-weighted) imposed on imports that compete with the output of the domestic sector.
    - Purpose: directly affects domestic demand for the sector’s output.
  - The downstream tariff:
    - Definition: the average cumulative tariff the sector’s output faces when exported either directly or indirectly through other (intermediate-output-using) sectors and countries.
    - Purpose: affects international demand for the sector’s output.
    - Calculation: analogous to upstream tariff, using a global input-output matrix scaled by the sector’s output.
  - (Note on methodology) Construction of upstream and downstream tariffs relies on Vandenbussche, Connell, and Simons (2017) for theoretical justification and follows Rouzet and Miroudot (2013) in practical implementation. Further details are provided in Online Annex 4.3.

*International Monetary Fund | April 2019*

### CHAPTER 4

### CHAPTER 4

### Empirical findings on tariffs and value chains
- Data and scope:
  - Large panel data set of 35 countries and 13 manufacturing sectors, controlling for country-specific macroeconomic changes and country-industry characteristics.
- Main empirical conclusions:
  - Tariffs up and down the value chain matter for output and productivity generally much more than domestic protection of the given sector.
  - Upstream and downstream tariffs have statistically significant negative effects on real value added, employment, labor productivity, and total factor productivity.
    - Mechanisms: upstream tariffs increase input costs; downstream tariffs reduce international demand for the sector’s output; both reduce returns to scale from international market participation.
  - Tariffs aimed at domestic protection do not appear to have significant effects except for a small negative effect on employment (possible offset between reduced competition and larger domestic market share).
  - Evidence of trade diversion:
    - The diversion tariff is positive and statistically significant for value added and employment, consistent with firms benefiting when competitors face tariffs.
    - The diversion tariff is insignificant for labor productivity and total factor productivity.
- Table 4.1 (sign and significance patterns summarized):
  - Real Value Added: Upstream Tariff (negative), Domestic Protection (positive), Downstream Tariff (negative), Diversion Tariff (positive).
  - Number Employees: Upstream Tariff (negative), Domestic Protection (negative), Downstream Tariff (negative), Diversion Tariff (positive).
  - Labor Productivity: Upstream Tariff (negative), Domestic Protection (insignificant), Downstream Tariff (negative), Diversion Tariff (insignificant).
  - Total Factor Productivity: Upstream Tariff (negative), Domestic Protection (positive), Downstream Tariff (negative), Diversion Tariff (insignificant).

### Illustrative simulations and amplification by global value chains
- Simulation scope and limitations:
  - Simulations illustrate partial equilibrium effects using coefficients from the empirical model; they do not capture channels such as policy uncertainty, confidence effects, or financial conditions; general equilibrium approaches are discussed elsewhere in the chapter.
  - 2011 is the most recent year for which the simulation could be carried out given data constraints; 2011 is taken as a good approximation of current global value chain links.
- Impact of greater integration:
  - Closer integration into global value chains has increased the sensitivity of upstream and downstream tariffs to nominal tariff changes, amplifying their effects.
  - A 1 percentage point tariff increase by all countries would have a larger negative effect in 2011 than in 1995.
  - The effect of a nominal tariff hike on real value added has become more negative for all countries, with larger overall GDP effects for countries with large and highly integrated manufacturing sectors (Germany, Korea, China, Japan) and smaller effects for countries with smaller, less integrated manufacturing sectors (Canada, United States).

### Bilateral and discriminatory tariff spillovers
- Patterns when tariffs are discriminatory (example: 1 percentage point reciprocal tariff increase between China and the United States):
  - China and the United States are the most affected and both are hurt by the move.
    - For China: downstream and, to a slightly lesser extent, upstream tariffs dominate.
    - For the United States: upstream tariffs are more important because intermediate imports from China play a relatively bigger role.
  - Third-country effects:
    - Trade diversion offsets negative spillovers for many third countries.
    - Mexico and Canada benefit most from substitution of China’s exports to the United States, reflecting proximity and strong trade relations with the United States.
    - East Asia also benefits to some extent.
    - Japan and Korea, which supply inputs to China, are affected by downstream tariffs but also benefit from trade diversion.
    - For Canada, trade diversion is relatively more important and the overall effect is most likely to be positive.
  - The smaller cost on the United States in percent of GDP reflects the relatively smaller weight of the US manufacturing sector in US GDP; the change in manufacturing value added itself is actually somewhat larger for the United States than for China.
- General equilibrium scenario: hypothetical large US–China trade dispute (three modeling approaches)
  - China and the United States suffer the largest losses from reciprocal tariff increases due to collapse in bilateral trade and lower returns to capital from tariff distortions.
  - Trade diversion leads to substitution of China’s exports to the United States toward Mexico, Canada, and east Asia.
  - No economically significant change in either country’s aggregate trade balance despite large bilateral trade reduction.
  - Macroeconomic spillovers in third countries are modest, but sectoral spillovers are larger as global value chains are repositioned.
  - Long-term sectoral shifts: sizable shifts in manufacturing capacity away from China (and the United States) toward Mexico, Canada, and east Asia, implying sizable job losses in specific sectors—especially in China and the United States.
  - Broader effects: a US–China trade war would also trigger increased uncertainty, negative confidence effects, and tightening of global financial conditions, with negative effects on most countries—so most countries are likely to be worse off even if some benefit from trade diversion.

### Gross versus value-added trade balances
- Rationale:
  - Gross bilateral balances (exports minus imports) can misrepresent the true origin and destination of value generated when production is carried out through global value chains and intermediate goods cross borders multiple times.
- Key observations:
  - Large gross bilateral imbalances are generally accompanied by large value-added imbalances, but value-added imbalances are typically smaller.
  - Example: United States, Korea, and China in electrical goods—gross deficits and surpluses can shrink when measured in value-added terms because intermediate inputs embody value from multiple countries.
  - Panel evidence for 2015:
    - The 10 largest bilateral imbalances in gross terms are generally smaller in value-added terms.
    - For the United States, trade deficits with Mexico and China shrink in value-added terms; the US surplus in services is reduced and its manufacturing deficit shrinks when measured in value-added terms.
  - Aggregate trade balance of a country is the same whether measured in gross or value-added terms when measured against the totality of the rest of the world.

### Conclusions and policy recommendations
- Conclusions:
  - Aggregate imbalances tend to be reflected at the bilateral level, while bilateral imbalances are of little consequence for the aggregate: attempts to influence a particular bilateral trade balance are likely to lead to compensating adjustments elsewhere, leaving the overall trade balance unchanged absent changes in macroeconomic conditions.
  - Over the past two decades, macroeconomic factors played a key role in explaining changes in bilateral balances (relative movement of partners’ domestic imbalance between supply and demand), reflecting fundamental factors, macroeconomic policies (such as fiscal policies and credit cycles), and, in some cases, exchange rate and supply-side policies.
  - Declines in tariffs and other trade costs enabled growth of global value chains, specialization by comparative advantage, and more efficient cross-border production arrangements.
  - Higher tariffs would have significant negative impacts on value added, employment, and productivity for the countries involved and for third countries through value chain links; greater international division of labor increases the scope for negative spillovers and spillbacks.
  - While some countries may benefit from trade diversion, escalation of trade tensions would raise uncertainty, lower confidence, and tighten global financial conditions, making most countries worse off.
- Two main policy conclusions:
  - Focus on aggregate trade balances and current accounts: Given the important role of macroeconomic imbalances in bilateral trade, discussion of external balances should focus on aggregate trade balances and current accounts—and the macroeconomic policies and distortions driving them. Aggregate external imbalances are not inherently bad, as they can finance investment and smooth consumption, but policymakers should avoid distortive macroeconomic policies that create excessive and possibly unsustainable imbalances.
  - Strong case for lowering tariffs: Lower tariffs will not only boost trade but also allow the international division of labor to better reflect comparative advantage, leading to output, employment, and productivity gains for countries up and down the value chain. Accompanying policies are important to ensure benefits from trade are widely shared and the burden of adjustment does not fall on only a few.

*Source: CHAPTER 4 — ThE DRIvERS OF BILaTERaL TRaDE aND ThE SPILLOvERS FROM TaRIFFS, International Monetary Fund | April 2019*

### Box 4.1 (continued)

### Box 4.1 (continued)

### Bilateral and aggregate trade balances: analytical relationship
- Under the assumption of symmetric trade costs, the gravity relationship can be rearranged to obtain:
  - TB_ij / (Y_i Y_j^W) = m_ij ∙ (TB_i / Y_i − TB_j / Y_j)
  - TB denotes trade balance; Y denotes output; Y_j^W is one output expressed as a share of world output; m_ij summarizes bilateral trade intensity.
- Implications:
  - The appropriately scaled bilateral trade balance depends on the relative evolution of each country’s aggregate trade balance-to-GDP ratio.
  - Double scaling (accounting for both countries’ output levels) captures the intuition that, all else equal, bilateral trade balances increase in absolute value as countries grow.
  - Changes in a country’s aggregate trade imbalance (driven by fundamentals such as fiscal policy and credit cycles) are amplified or dampened at the bilateral level by m_ij, which reflects pair-specific bilateral and multilateral trading costs and micro-structural determinants.

### US–China bilateral trade balance: empirical illustration
- When scaled by US GDP alone, the US–China bilateral trade imbalance did not shrink after the 2008–09 global financial crisis; when double scaling (including China’s growing world share) is applied, a notable reduction emerges.
- Empirical fit:
  - A “predicted” scaled bilateral trade balance constructed as (constant m_ij) × (difference in aggregate trade-balance-to-GDP ratios) tracks the actual scaled bilateral balance closely, indicating aggregate trade-balance variations explain the evolution of the (scaled) US–China bilateral balance well.
  - Deviations reflect changes in trade intensity (m_ij), e.g., evolving world trade costs.
- Figure labels/notes:
  - Figure 4.2.1 panels show Bilateral TB/GDP_USA and Bilateral TB/(GDP_USA * GDP_CHN share) (right scale).
  - Note: TB = Trade Balance. Data labels use ISO country codes. Scaling note: As a percentage of US GDP times the world share of China’s GDP.

*The author of this box is Roberto Piazza.*

### Do bilateral-targeted policies reduce overall current account deficits? (Box 4.2 & Box 4.3 content)
- Research question: Will a policy of targeting bilateral trade deficits reduce a country’s overall current account deficit?
- Identification of large trade deficit reversal episodes (following Milesi-Ferretti and Razin 1998):
  - Criteria: (1) average reduction in overall trade deficit ≥ 3 percentage points of GDP over three years relative to the three-year average before the event; (2) maximum trade deficit in the three years after reversal < minimum deficit in the three years preceding the reversal; (3) no other episode in the following six years.
  - Using IMF Direction of Trade Statistics data from 1980–2017 for countries with nominal GDP above the world median in 2017 (excluding fuel exporters), 92 large deficit-adjustment episodes were identified.
- Key empirical findings:
  - Only 17 percent (16 out of 92) of large overall deficit-adjustment episodes were associated with large bilateral trade adjustment by at least one of the five biggest deficit partners.
  - Robustness: lowering the bilateral adjustment threshold to 2½ percentage points increases the number of episodes with large bilateral adjustments from 16 to 22.
  - Overall trade adjustments are not generally driven by large adjustments in a country’s top trading partners.
  - Large adjustments in at least one of the five biggest deficit partners sometimes occur without a large reversal in the overall trade deficit.
- Distribution of adjustments:
  - Absolute adjustments concentrated at the top:
    - The five biggest deficit partners are, on average, responsible for 54 percent of deficit correction.
    - The next five are responsible for 12 percent of the correction.
    - The following five explain 8 percent of the correction.
  - Relative adjustments (adjustment relative to initial bilateral deficit) are more evenly distributed:
    - Mean adjustment for the top five deficit partners is about 48 percent of their initial level.
    - Mean adjustment for deficit partners ranked six to ten is about 52 percent of their initial level.
  - Regression analysis: all trading partners (top or bottom) contribute to trade deficit adjustments; disproportionate reductions with top trading partners are not necessary for overall trade balance reduction.
- Conclusion:
  - Targeting bilateral trade balances would likely not help reduce a country’s overall current account deficit.
  - This is consistent with conventional wisdom that changes in current account balances are best achieved through adjustments to macroeconomic policies (saving and investment), not trade policies.

*The authors of this box are Kyun Suk Chang, Swarnali Ahmed Hannan, and Sergii Meleshchuk.*

### Macro and micro effects of a hypothetical 25 percentage point increase in US–China tariffs (Box 4.4)
- Scenario: tariffs on all US–China goods trade increase by 25 percentage points.
- Models used:
  - GIMF: Global Integrated Monetary and Fiscal model (dynamic stochastic general equilibrium).
  - GTAP: Global Trade Analysis Project (multisector perfect-competition computable general equilibrium).
  - CFRT: multisector heterogenous-firm model with entry and exit à la Melitz (Caliendo and others 2017).
- Model emphases:
  - GIMF: aggregate effects over time; short term amplified by nominal and real rigidities; medium/long term effects through permanent impacts on capital and labor returns.
  - GTAP and CFRT: sectoral/regional reallocation effects; CFRT amplifies reallocation via increasing returns to scale and firm-level fixed costs.
- Calibration note:
  - Substitution elasticities calibrated so substitution between any two foreign suppliers is easier than between a foreign supplier and a domestic firm; this amplifies positive effects on third countries by diverting production toward them rather than re-shoring.

Effects on US and China
- Short- and long-term trade collapse:
  - US–China trade falls by 25–30 percent in the short term (GIMF).
  - US–China trade falls somewhere between 30 percent and 70 percent over the long term, depending on model and direction of trade.
- Annual real GDP losses:
  - United States: range from –0.3 percent to –0.6 percent.
  - China: range from –0.5 percent to –1.5 percent.
- Asymmetry and terms of trade:
  - Effects on China typically larger because exports to the United States represent a larger share of the Chinese economy.
  - In GIMF, China’s short-term effects are more pronounced due to insufficient wage and price adjustment.
  - In CFRT, China’s losses are amplified by loss of economies of scale.
  - Terms of trade improve in the United States and worsen in China.
- Multilateral trade balance:
  - Although the US–China bilateral trade deficit is reduced, there is no economically significant change in each country’s multilateral trade balance (GIMF result).

Macro spillovers to third countries
- Robust cross-model result:
  - Third countries increase exports to the United States and decrease exports to China.
  - Much of global trade diversion stems from third countries expanding exports to the United States at the expense of China, while importing more intermediate goods from China.
- Third-country export effects:
  - Generally positive for third-country overall exports, with Mexico and Canada benefiting the most due to proximity and strong US trade relations.
  - Model variation:
    - GIMF: largest increase in third-country exports in the short term.
    - CFRT: sizable positive effects for some countries.
    - GTAP: weaker and in some cases negative responses.
- Output in third countries:
  - Effects vary across models; typically modest in size.
  - GIMF shows positive short-term effects in other North America (excluding the United States) due to trade diversion and nominal rigidities.
  - Long run: small negative effects in most regions as capital-reducing distortions from higher tariffs spill over to third countries, except other North America (excluding the United States).
  - GTAP: positive but small effects.
  - CFRT: some countries (Mexico, Canada, Malaysia, Thailand, Korea) experience relatively large output expansions benefiting from economies of scale.

Table highlights (GIMF / GTAP / CFRT results for selected bilateral flows)
- Trade Balance (percent of GDP), GIMF:
  - China: –0.2 (Year 1), –0.3 (Long Run).
  - United States: –0.2 (Year 1), –0.3 (Long Run).
- Exports to the United States (selected rows from Table 4.4.1):
  - China: –20.9–25.1–71.3–56.0
  - Asia: 3.7 1.8 10.6 7.7
  - Euro Area: 3.8 1.7 8.4 5.7
  - NAFTA: 2.6 0.7 3.2 3.2
  - Rest of the World: 3.1 1.7 6.1 5.8
- Exports to China (selected rows from Table 4.4.1):
  - United States: –27.5–36.4–77.6–63.7
  - Asia: 0.5 –1.5 –4.9 –1.1
  - Euro Area: 1.3 –1.6 –4.3 –1.4
  - NAFTA: –0.4 –2.6 –4.4 –2.9
  - Rest of the World: –0.1 –1.6 –1.0 –2.8

Sectoral reallocations and labor market effects
- Aggregate spillovers moderate, but sectoral effects large and uneven:
  - Manufacturing shows a large worldwide contraction, with major fallout in Chinese electronics and other manufacturing sectors.
  - Manufacturing expands in Mexico, Canada, and some Asian countries.
  - Services expand in China and contract in several other countries.
  - US agricultural sector experiences a sizable contraction.
- Job losses (CFRT illustrative magnitudes):
  - US: about 1 percent of the workforce lost in the agricultural and transportation equipment sectors.
  - China: about 5 percent of the workforce lost in the other manufacturing sector.
- Magnitude varies across models; larger in CFRT due to economies of scale.

Repositioning of global value chains (electronics and machinery example)
- Electronics and machinery account for about 20 percent of world imports.
- CFRT results:
  - China would eventually stop being the number one exporter of electronics and machinery to the United States; other Asian countries, Canada, and Mexico replace China.
  - US imports of electronics and machinery: China before tariffs #1, 22.1%; after tariffs #4, 11.5%. East Asia moves to #1, 17.7%; Mexico moves from #3, 12.6% to #2, 14.6%; Canada moves from #4, 10.8% to #3, 12.3%.
  - Mexico: sizable entry of new firms into electronics leads to large increases in imports of intermediates from everywhere, especially from China.
- Figure 4.4.4 (Mexico’s imports of intermediate inputs for electronics and manufacturing):
  - Change in imports from China and change in imports from all partners by sector (agriculture, other manufacturing, transport, metals, oil and chemicals, wood and paper, textile, food, mining) showing shifts in intermediate sourcing patterns.

*The authors of this box are Carlos Caceres, Diego Cerdeiro, Rui Mano, Rafael Portillo, and Marika Santoro.*

### Box 4.4 (continued)

### Box 4.4 (continued)

### References
- Ahn, JaeBin, Emine Boz, Mauric Obstfeld, and Petia Topalova. Forthcoming. “Bilateral Trade Balances and Global Value Chains.” International Monetary Fund, Washington, DC.
- Ahn, JaeBin, Era Dabla-Norris, Romain Duval, Bingjie Hu, and Lamin Njie. 2016. “Reassessing the Productivity Gains from Trade Liberalization.” IMF Working Paper 16/77, International Monetary Fund, Washington, DC.
- Amiti, Mary, Mi Dai, Robert C. Feenstra, and John Romalis. 2017. “How Did China’s WTO Entry Affect US Prices?” NBER Working Paper 23487, National Bureau of Economic Research, Cambridge, MA.
- Amiti, Mary, and Jozef Konings. 2007. “Trade Liberalization, Intermediate Inputs, and Productivity: Evidence from Indonesia.” American Economic Review 97 (5): 1611–638.
- Anderson, James E., and Eric van Wincoop. 2003. “Gravity with Gravitas: A Solution to the Border Puzzle.” American Economic Review 93 (1): 170–92.
- Bacchetta, Marc, Cosimo Beverelli, Olivier Cadot, Marco Fugazza, Jean-Marie Grether, Matthias Helble, Alessandro Nicita, and Roberta Piermartini. 2012. “A Practical Guide to Trade Policy Analysis.” World Trade Organization, Geneva.
- Baldwin, Richard, and Javier Lopez-Gonzalez. 2013. “Supply-Chain Trade: A Portrait of Global Patterns and Several Testable Hypotheses.” NBER Working Papers 18957, National Bureau of Economic Research, Cambridge, MA.
- Baldwin, Richard, and Daria Taglioni. 2011. “Gravity Chains: Estimating Bilateral Trade Flows When Parts and Components Trade is Important.” Journal of Banking and Financial Economics 2 (2): 61–82.
- Barattieri, Alessandro. 2014. “Comparative Advantage, Service Trade, and Global Imbalances.” Journal of International Economics 92 (1): 1–13.
- Boz, Emine, Nan Li, and Hongrui Zhang. 2019. “Effective Trade Costs and the Current Account: An Empirical Analysis.” IMF Working Paper 19/8, International Monetary Fund, Washington, DC.
- Caliendo, Lorenzo, Robert C. Feenstra, John Romalis, and Alan Taylor. 2017. “Tariff Reductions, Entry, and Welfare: Theory and Evidence for the Last Two Decades.” NBER Working Paper 21768, National Bureau of Economic Research, Cambridge, MA.
- Caliendo, Lorenzo, and Fernando Parro. 2015. “Estimates of the Trade and Welfare Effects of NAFTA.” The Review of Economic Studies 82 (1): 1–44.
- Criscuolo, Chiara, and Jonathan Timmis. 2017. “The Relationship between Global Value Chains and Productivity.” Centre for the Study of Living Standards, International Productivity Monitor 32 (Spring): 61–83.
- Daruich, Diego, William Easterly, and Ariell Reshef. 2019. “The Surprising Instability of Export Specializations.” Journal of Development Economics 137: 36–65.
- Davis, Donald, and David Weinstein. 2002. “The Mystery of the Excess Trade (Balances).” American Economic Review 92 (2): 170–74.
- Feenstra, Robert C. 2004. Advanced International Trade: Theory and Evidence. Princeton, NJ, and Oxford: Princeton University Press.
- Feenstra, Robert C., Philip Luck, Maurice Obstfeld, and Katheryn Russ. 2018. “In Search of the Armington Elasticity.” Review of Economics and Statistics 100 (1): 135–50.
- Furceri, Davide, Swarnali A. Hannan, Jonathan D. Ostry, and Andrew K. Rose. 2018. “Macroeconomic Consequences of Tariffs.” CEPR Discussion Papers 13389, Center for Economic and Policy Research, Washington, DC.
- Group of Twenty (G20). 2018. “G20 Global Forum on Steel Excess Capacity Ministerial Meeting, Ministerial Report.” September 20, Paris.
- Head, Keith, and Thierry Mayer. 2014. “Gravity Equations: Workhorse, Toolkit, and Cookbook.” In Handbook of International Economics, Volume 4, edited by Gita Gopinath, Elhanan Helpman, and Kenneth Rogoff. Oxford: Elsevier.
- International Monetary Fund (IMF). 2011. “People’s Republic of China: Spillover Report for the 2011 Article IV Consultation and Selected Issues.” IMF Country Report 11/193, Washington, DC.
- International Monetary Fund (IMF). 2017a. “Fostering Inclusive Growth.” Prepared for discussion at the July 2017 Group of Twenty Leaders’ Summit.
- International Monetary Fund (IMF). 2017b. “People’s Republic of China 2017 Article IV Consultation—Press Release; Staff Report; and Statement by the Executive Director for the People’s Republic of China.” IMF Country Report 17/247, Washington, DC.
- International Monetary Fund, World Bank, and World Trade Organization (IMF/WB/WTO). 2017. “Making Trade an Engine of Growth for All: The Case for Trade and for Policies to Facilitate Adjustment.” Prepared for discussion at the March 2017 meeting of Group of Twenty 20 Sherpas by staffs of the International Monetary Fund, World Bank, and World Trade Organization.
- International Monetary Fund, World Bank, and World Trade Organization (IMF/WB/WTO). 2018. “Reinvigorating Trade and Inclusive Growth.” Prepared for discussion at the March 2017 meeting of the Group of Twenty Sherpas by staffs of the International Monetary Fund, World Bank, and World Trade Organization.
- Joy, Mark, Noëmie Lisack, Simon Lloyd, Dennis Reinhardt, Rana Sajedi, and Simon Whitaker. 2018. “Mind the (Current Account) Gap.” Bank of England Financial Stability Paper 43, London.
- Johnson, Robert C., and Guillermo Noguera. 2012a. “Accounting for Intermediates: Production Sharing and Trade in Value Added.” Journal of International Economics 86 (2): 224–36.
- Johnson, Robert C., and Guillermo Noguera. 2012b. “Proximity and Production Fragmentation.” American Economic Review Papers and Proceedings 102 (3): 407–11.
- Johnson, Robert C., and Guillermo Noguera. 2017. “A Portrait of Trade in Value Added over Four Decades.” Review of Economics and Statistics 99 (5): 896–911.
- Koopman, Robert, Zhi Wang, and Shang-Jin Wei. (2014). Tracing Value Added and Double Counting in Gross Exports. American Economic Review 104 (2): 459–94.
- Milesi-Ferretti, Gian Maria, and Assaf Razin. 1998. “Sharp Reductions in Current Account Deficits: An Empirical Analysis.” European Economic Review 42 (3–5): 897–908.
- Rouzet, Dorothée, and Sébastien Miroudot. 2013. “The Cumulative Impact of Trade Barriers along the Value Chain: An Empirical Assessment Using the OECD Inter-Country Input-Output Model.” Organisation for Economic Co-operation and Development, Paris.
- Silva, JMC Santos, and Silvana Tenreyro. 2006. “The Log of Gravity.” The Review of Economics and Statistics 88 (4): 641–58.
- Topalova, Petia, and Amit Khandelwal. 2011. “Trade Liberalization and Firm Productivity: The Case of India.” The Review of Economics and Statistics 93 (3): 995–1009.
- Vandenbussche, Hylke, William Connell, and Woutel Simons. 2017. “Global Value Chains, Trade Shocks, and Jobs: An Application to Brexit.” CEPR Discussion Paper 12303, Centre for Economic Policy Research, London.
- World Bank (WB), World Trade Organization, Organisation for Economic Co-operation and Development, and University of International Business and Economics. 2017. Global Value Chain Development Report 2017: Measuring and Analyzing the Impact of GVCs on Economic Development. Washington, DC: World Bank.
- Yotov, Yoto V., Roberta Piermartini, José-Antonio Monteiro, and Mario Larch. 2016. “An Advanced Guide to Trade Policy Analysis: The Structural Gravity Model.” World Trade Organization, Geneva.

*World Economic Outlook: Growth Slowdown, Precarious Recovery, International Monetary Fund | April 2019*

---


_Source: https://www.imf.org/-/media/files/publications/weo/2019/april/english/ch4.pdf_