## wp1746

## Source details

**Canonical URL:** [wp1746](https://www.imf.org/-/media/files/publications/wp/2017/wp1746.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2017/wp1746.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2017/wp1746.pdf.json)

---

### Introduction and headline findings
- Paper title: Revisiting the Link between Trade, Growth and Inequality: Lessons for Latin America and the Caribbean.
- Authors: Kimberly Beaton, Aliona Cebotari and Andras Komaromi. Authorized for distribution by Valerie Cerra. March 2017.
- Approaches used:
  - Cross-country panel regressions.
  - Event studies of trade liberalization episodes.
- Core messages:
  - Trade openness and connectivity to the center of the trade network has substantial macroeconomic benefits.
  - No statistically significant or economically sizable direct negative impact of trade on overall income inequality was found.
- Regional context and salient numeric facts:
  - The region accounts for about 8 percent of global economic activity and about 5 percent of global exports of goods and services.
  - LAC’s trade openness accounts for only about 44 percent of regional GDP.
  - WTO accession of China occurred in 2001 (noted as a landmark in global trade integration).

---

### Methodology and identification
- Empirical strategies:
  - Cross-country panel regressions using dynamic growth models and system GMM (Arellano and Bover (1995); Blundell and Bond (1998)).
  - Event studies using Sachs and Warner (1995) dates updated by Wacziarg and Welch (2003, 2008), focusing on SWWW liberalization dates after 1970 (74 episodes between 1970 and 2001).
- Growth regression setup:
  - Dependent variable: average rate of real per capita GDP growth measured using non-overlapping 5-year averages.
  - Controls include human and physical capital, FDI as a share of GDP, and the terms of trade.
  - Benchmark trade measure: trade openness (OP); augmented specification adds trade characteristics TC (connectivity, GVC participation, export market and product characteristics).
- System GMM specifics:
  - Estimates a system of two linear IV regressions (levels and differences).
  - Internal instruments: lagged values of dependent and explanatory variables; instruments for levels are lagged differences, instruments for differences are lagged levels.
  - Instrument set restricted to avoid overfitting: only the first and second appropriate lag of each explanatory variable; for period averages instruments correspond to the average of periods t-2 and t-3; for initial values only the first appropriate lag (observation at start of period t-1).
- Specification tests and robustness:
  - Hansen test of over-identifying restrictions.
  - Test of second-order serial correlation in the first differenced error term (AR(2)).
  - Alternative instrument sets tested and generally confirm baseline results.
- Sample sizes:
  - Trade-growth regressions benchmark: 131 countries in the benchmark specification; unbalanced panel.
  - Inequality regressions: 118 countries; five-year panels over 1980–2013.
  - Event studies: 74 liberalization episodes (1970–2001).

---

### Trade and economic performance — main regression findings
- Trade openness and growth:
  - Benchmark growth regression (Table 1) finds trade openness has a positive and statistically significant impact on average per capita economic growth.
  - Evidence of conditional convergence: Initial GDP per capita coefficients reported (examples): -3.816***, -8.825***, -3.713***, -4.187*** (standard errors: (0.759), (2.961), (0.718), (0.907)).
  - Trade openness coefficients reported (examples): 2.045***, -0.0408, 1.658***, 0.426 (standard errors: (0.574), (1.494), (0.574), (1.298)).
- Other growth determinants (selected):
  - Public Infrastructure coefficients: 1.674***, 5.094***, 1.895***, 0.643 (standard errors: (0.407), (1.813), (0.394), (1.019)).
  - Terms of trade have an adverse effect on growth in several specifications (examples: -1.741, 2.591, -1.495, -1.604 with reported standard errors).
  - FDI (percent of GDP) coefficients: 0.274, 0.430, 0.362*, 0.399 (standard errors: (0.238), (0.533), (0.217), (0.266)).
- Tests and instrument validity (examples):
  - Observations: 1,041 in each column (Table 1).
  - Number of ISO: 135.
  - AR(1): 0.0190, 0.0299, 0.0190, 0.0192.
  - AR(2): 0.244, 0.309, 0.254, 0.241.
  - Hansen: 0.5110, 0.006940, 0.9010, 0.577.

---

### Heterogeneity by development level and region
- Advanced-economy interaction:
  - Advanced economy dummy interacted with trade openness: Advanced economy (dummy)*Trade Openness: 10.97** (standard error: (5.551)); Advanced economy (dummy): -38.05* (standard error: (22.07)).
  - Suggests impact of trade openness on growth is stronger for advanced countries in some specifications; Hansen test suggests weak instruments in that regression, so interpret with caution.
- LAC interaction:
  - Latin America and the Caribbean (dummy)*Trade Openness: 6.723** (standard error: (2.968)); Latin America and the Caribbean (dummy): -2.196*** (standard error: (0.814)).
  - Suggests a stronger economic impact of trade for LAC relative to other regions in some specifications.
- Regional contrast:
  - Converse (not reported) result for an Asian dummy suggests a lower impact for Asian economies, supporting the relevance of trade-connection characteristics.

---

### Trade connectivity and network measures — growth effects
- Network indicators capture granular aspects of integration beyond trade openness: total degree, out-degree, in-degree, eigencentrality, betweenness, closeness, hub/authority scores.
- Key growth effects (Table 2 and text):
  - Moving from the 25th percentile to the 75th percentile of total trade partners could improve average growth outcomes by about 3 percentage points over a five-year period; corresponds to an increase of 142 trading partners.
  - Increasing export markets (out degree) by 85 partners (25th to 75th percentile) would enhance average growth by 1.8 percentage points over a 5-year period.
  - Number of export markets (Out Degree) is significant and positive for growth; number of import source countries (In Degree) is insignificant.
  - Eigencentrality is positively associated with growth; centrality among importers or exporters separately is less important than overall centrality.
- Selected coefficients (Table 2, system GMM):
  - Number of trading partners (Total degree): 7.695***.
  - Out Degree: 3.846*.
  - In Degree: 3.992 (insignificant).
  - Eigencentrality: 9.787***.
  - Observations: 829; Number of ISO: 131.
  - AR(1): 0.0595; AR(2): 0.390; Hansen: 0.199.

---

### Global Value Chains (GVCs): participation and growth
- GVC participation measures:
  - Sum of foreign value added in gross exports and indirect value added (Eora MRIO and Koopman et al. decomposition).
  - Foreign value added (FVA/gross exports) indicates downstream (backward) orientation; indirect value added indicates upstream (forward) orientation.
- Growth effects:
  - Increasing GVC participation from the 25th to the 75th percentile could enhance 5-year average per capita growth by 1.8 percentage points.
  - Upstream participation (indirect value added) has the more important effect on growth; downstream participation has a positive but not statistically significant effect.
  - Domestic value added in intermediate goods exports re-exported to third countries/gross exports coefficient: 0.485*** (Table 3, column (3)).
- Selected GVC-related coefficients (Table 3):
  - Global Value Chain Participation 2/: 0.0909**.
  - Indirect Value Added 3/: 0.0889*.
  - FVA/Gross Exports: 0.0304; 0.0657 (columns reported).
  - Observations: 652; Number of ISO: 135.
  - AR(1): 0.0197; AR(2): 0.582; Hansen: 0.0666.
  - Caution: Hansen test results indicative of weak instruments for regressions with GVC measures; interpret with caution.
- LAC patterns:
  - LAC’s participation in GVCs remains limited relative to Asia.
  - Mexico and Central America are more downstream/final-stage participants; South American countries are more upstream/earlier-stage participants.
  - LAC countries tend to participate in relatively shorter GVCs.
  - Bringing LAC’s participation in GVCs in line with Asia could increase the region’s real per capita growth by about 1 percentage point.

---

### Export markets, product characteristics, and growth
- Export markets:
  - Trade connections with regional partners associated with larger growth effects than inter-regional trade: Regional partners coefficient in Table 4: 1.019*** (0.369).
  - LAC: about a quarter of total exports destined to markets within the region; many LAC countries have regional export shares well below the 50th percentile.
  - Increasing intra-regional trade in LAC from about 15 percent of the region’s exports to 55 percent could enhance regional per capita growth by 1.3 percentage points.
  - Share of emerging markets and developing countries in LAC countries’ total exports increased about 50 percent, from 33 percent on average over 1995-1999 to about 50 percent on average over 2010-2015.
  - Trade with the most central importers associated with stronger growth: Most central importers coefficient in Table 4: 0.549* (0.293); increasing share of trade with most central importers from the 25th percentile to the 75th percentile would increase average per capita growth by 0.8 percentage points.
  - Number of trade agreements positively associated with growth: coefficient 2.296*** (0.544); increasing number of trade agreements from the 25th percentile to the 75th percentile associated with a 1.5 percentage point increase in average per capita growth.
- Export product characteristics:
  - Higher export quality linked to stronger growth effects (Table 5, column 1).
  - Increasing diversification of LAC’s export product basket by 25 percent has the potential to increase the region’s real per capita growth by about 1 percentage point.
  - LAC has revealed comparative advantage in mineral fuels and non-fuel primary commodities and lags in skill- and technology-intensive manufactured products.
  - Economic complexity has a positive, but not significant, effect on growth in reported regressions.
- Measurement notes:
  - Export diversification measured using the IMF Export Diversification Index (higher values indicate lower diversification / higher concentration).
  - Export quality measured by the IMF Export Quality Index (higher values indicate higher export quality).
  - Economic complexity measured using the Observatory of Economic Complexity.

---

### Event studies: liberalization episodes — growth, volatility, and structural change
- Main stylized outcomes from event studies (74 episodes, 1970–2001):
  - Growth and investment pick up after liberalization; effects persist for several years.
  - FDI inflows tend to increase following liberalization.
  - Economic volatility tends to decline after liberalization.
  - Reforms often preceded by low-growth and high-volatility episodes.
- Structural changes:
  - Manufacturing export composition shifts toward medium skill-intensity with declining share of high-skill goods.
  - Production structure: decline in agriculture share of value added and increase in tertiary (services) share.
  - Structural changes imply winners and losers; potential significant adjustment costs for some segments (e.g., rural areas and some manufacturing industries).
- Selected event-study and panel-regression coefficients (examples reported):
  - Liberalization indicator effects (Table 6, selected): Per capita growth: 2.663***; Investment rate: 2.307**; Services share: 3.432*; High-skill manufacturing share: -0.122***; Medium-skill manufacturing share: 0.082***.
  - Panel-regression examples: Initial GDP per capita: -3.822***, -2.759***, -4.047***, -3.740***; Public Infrastructure: 1.417***, 1.335***, 1.467**, 1.552**; Trade openness: 1.860**, 0.0791, 1.705**, 1.223; FDI (percent of GDP): 0.229, 0.507***, 0.294, 0.327.
  - Observations and sample sizes vary by specification (examples): Observations: 989, 784, 995, 969; Number of ISO: 133, 110, 129, 129; AR(1): 0.0271, 0.062, 0.0219, 0.0245; AR(2): 0.386, 0.353, 0.238, 0.240; Hansen: 0.233, 0.648, 0.219, 0.546.

---

### Trade and inequality — regression and event-study evidence
- Aggregate relationship:
  - Cross-country panel regressions and event studies find no robust evidence that trade openness increases overall income inequality.
  - Panel fixed effects regressions (five-year panels over 1980–2013) generally report trade openness coefficients on Gini that are negative but not statistically significant in most specifications.
- Selected inequality-regression coefficients (examples):
  - Trade Openness: -0.0205, -0.00851, -0.0413*, -0.0216, -0.0220, -0.00796 (various specifications).
  - Financial Deepening: 0.0342*** (Market Gini) and 0.0215** (Net Gini) in one reported specification.
  - Financial Openness: small positive but not significant coefficients in reported specifications.
  - Constants: 48.74***, 39.67***, 48.12***, 39.28***, 48.80***, 39.65***.
  - Observations reported in some inequality regressions: 562; Number of ISO: 118; R-squared examples: 0.133, 0.120, 0.143, 0.126.
- Event-study inequality evidence:
  - Newly liberalizing countries did not experience a larger increase in income inequality than the world average; in some comparisons they experienced a smaller increase.
  - For LAC (21 episodes), the average liberalizing country in LAC fared better than the rest of the world in changes in Gini.
  - Simple panel regressions: liberalization indicator coefficient is insignificant for both market and net Gini (Table 8 summary: liberalization coefficient examples: -0.447, 0.06; Observations: 3557; Number of countries: 138; Adjusted R-squared (within): 0.198 (Market Gini), 0.14 (Net Gini)).
- Skill-premium evidence (SEDLAC five countries: Argentina, Brazil, Panama, Uruguay, Venezuela):
  - No clear trend in skill premiums around liberalization episodes; skill premium measures show considerable variability and almost no systematic increase after liberalization.

---

### Policy recommendations and implications
- Trade and growth policy priorities:
  - Enhance trade integration inter- and intra-regionally to maximize growth benefits.
  - Prioritize regional integration as a pathway to enhance growth dividends of trade in an environment of rising global protectionism.
  - Promote participation in global value chains (GVCs) via:
    - Infrastructure improvements.
    - Strengthening human capital.
    - Supportive policies for research and development to capture knowledge and technology spillovers.
  - Support export quality improvements and export diversification.
  - Remove trade barriers, with emphasis on the import of intermediate goods because barriers compound across multiple border crossings in GVCs.
  - Expand and deepen formal trade agreements (including broader coverage beyond tariffs) to increase growth; number of trading partners with a trade agreement coefficient reported as 2.296*** (0.544).
- Inclusion and adjustment:
  - Trade can promote growth without adversely affecting overall income inequality, but trade integration can impose adjustment costs on particular segments of society.
  - Implement targeted social safety nets to support adjustment.
  - Address the equity gap in redistributive policies in LAC given the disparity between market and net Gini coefficients; policy actions include overhauling regressive tax systems, improving tax administration efficiency, and improving targeting of social programs.

---

*Source: WP/17/46 (wp1746), Sections 1–6 (IMF Working Paper, March 2017).*

### Section 1

### wp1746 - Section 1

### Introduction: scope and headline findings
- Paper title: Revisiting the Link between Trade, Growth and Inequality: Lessons for Latin America and the Caribbean.
- Authors: Kimberly Beaton, Aliona Cebotari and Andras Komaromi. Authorized for distribution by Valerie Cerra. March 2017.
- Two complementary approaches are used: cross-country panel regressions and event studies of trade liberalization episodes.
- Two broad, consistent messages from both approaches:
  - Trade openness and connectivity to the center of the trade network has substantial macroeconomic benefits.
  - No statistically significant or economically sizable direct negative impact of trade on overall income inequality was found.
- Regional context and salient numeric facts:
  - The region accounts for about 8 percent of global economic activity and about 5 percent of global exports of goods and services.
  - LAC’s trade openness accounts for only about 44 percent of regional GDP.
  - WTO accession of China occurred in 2001 (noted in narrative as a landmark in global trade integration).
- Policy implications emphasized:
  - Enhancing trade integration inter- and intra-regionally to maximize growth benefits.
  - Promote participation in global value chains (GVCs) via infrastructure improvements, strengthening human capital, and supportive policies for research and development to capture knowledge and technology spillovers.
  - Support export quality improvements and export diversification.
  - In the context of rising protectionist sentiment globally, regional integration may currently present the greatest opportunity for the region to strengthen trade dividends.

### Methodology and data: empirical strategy and identification
- Two empirical approaches:
  - Cross-country panel regressions: leverage broad country coverage and long time periods while controlling for non-trade determinants and unobserved country-specific factors.
  - Event studies of major trade liberalization episodes: natural experiments focusing on episodes with large trade policy changes; useful robustness check but subject to omitted variable caveat when trade reforms are part of broader policy packages.
- Main econometric framework (growth regressions):
  - Benchmark: standard growth model with log of GDP per capita as dependent variable and a set of control variables (CV), trade openness (OP), unobserved time-specific effects (υt), unobserved country-specific effects (μi), and error term (εit).
  - Control variables include human and physical capital, foreign direct investment as a share of GDP, and the terms of trade.
  - Trade openness used as the benchmark measure of trade integration; event studies serve as complementary robustness.
- Augmented specification:
  - Adds a set TC of trade characteristics (connectivity, GVC participation, export market and product characteristics) to assess how trade characteristics influence growth outcomes while controlling for the level of trade integration.
- Dynamics and estimation:
  - Dependent variable in trade-growth regressions is the average rate of real per capita GDP growth measured using non-overlapping 5-year averages.
  - Regressions are dynamic and include the level of per capita GDP at the start of the corresponding period to capture conditional convergence effects.
  - Estimation method: system generalized method of moments (SGMM) as in Arellano and Bover (1995) and Blundell and Bond (1998).
  - SGMM strategy specifics:
    - Estimates a system of two linear IV regressions: one in levels and one in differences.
    - Internal instruments are used (lagged values of dependent and explanatory variables) to address dynamic panel bias and endogeneity.
    - Instruments for the levels equation are lagged differences; instruments for the differenced equation are lagged levels.
- Instrument selection and overfitting avoidance:
  - To avoid overfitting bias, the instrument set is restricted: only the first and second appropriate lag of each explanatory variable are used as instruments.
  - For variables measured as period averages, instruments correspond to the average of periods t-2 and t-3.
  - For variables measured as initial values, only the first appropriate lag is used, corresponding to the observation at the start of period t-1.

### Identification caveats and complementary strengths
- Endogeneity concerns are central: trade and growth are likely jointly endogenous.
- SGMM mitigates dynamic panel bias and endogeneity using internal instruments under the assumption that shocks to growth are unpredictable given past values of explanatory variables.
- Event studies mitigate endogeneity by focusing on large, discrete trade-policy changes, but cannot fully isolate trade effects from concurrently implemented policies (omitted variable problem).
- Using both methods provides cross-validation: consistent results across methodologies strengthen inference.

*Source: WP/17/46, Section 1 (IMF Working Paper, March 2017).*

### Section 2

### wp1746 - Section 2

### Methodology: Trade-Growth Regressions and Instrumentation
- Estimations of equation 2 include the characteristics of trade connections one at a time to avoid overextending the number of required instruments (following de la Torre et al (2015)).
- Specification tests used:
  - Hansen test of over-identifying restrictions to assess validity of the instrument set.
  - Test of second-order serial correlation in the first differenced error term; first-order serial correlation is expected even if the original error term (in levels) is uncorrelated. Second-order serial correlation would reject the appropriateness of the proposed instruments.
- Robustness checks: alternative instrument sets tested (estimates not reported) and generally confirm baseline results.
- Sample and data:
  - Trade-growth regressions estimated for a large unbalanced panel dataset.
  - 131 countries are included in the benchmark specification.
  - Annex 1 provides detail on the data used for each variable in the regressions.

### Inequality Regressions: Specifications and Controls
- Core specification (equation 3) links movements in countries’ Gini coefficients (market and net Gini) to explanatory variables:
  - The dependent variable is ܫܰܫܩ
௜,௧ (market or net Gini).
  - Vector of control variables, CV, includes measures of other aspects of globalization, education, government redistributive policy, shares of agriculture and services in value-added, and financial globalization/openness considerations.
  - ߤ
௧ are unobserved time-specific effects; ߛ
௜ are unobserved country-specific effects; ܳ
௜,௧ is the error term.
- Interaction specification (equation 4) adds interaction terms to test whether the impact of trade openness on inequality varies by development or location:
  - ܺ
௜,௧ is either a dummy variable for advanced economies or for LAC depending on the specification.
  - Coefficient ߚ
ஹ on the interaction term denotes the marginal impact of development level or location on the impact of trade openness on inequality.
- Estimation details:
  - Equations 3 and 4 estimated using panel fixed effects with five-year panels over 1980-2013.
  - Robust standard errors clustered at the country level.
  - 118 countries are included in these regressions.
  - Annex 1 provides detail on each variable in the regression.
- Note on labor market flexibility: a World Economic Forum–based measure of labor market regulation was considered but found insignificant and omitted from the final specification.

### Event Studies: Trade Liberalization Episodes and Identification
- Rationale: Studying large, discrete trade liberalization episodes provides an alternative way to gauge effects of trade openness on outcomes.
- Data sources and choices:
  - Rely on Sachs and Warner (1995) liberalization dates, updated and extended by Wacziarg and Welch (2003, 2008).
  - Focus on SWWW liberalization dates after 1970 due to data limitations.
  - 74 trade liberalization episodes between 1970 and 2001 are used to trace the average effect of trade opening.
- Empirical approach:
  - Two steps:
    1. Examine time path of economic performance, structure, and inequality for an average country before and after liberalization and compare to the world average.
    2. Run regressions of outcomes on a binary liberalization indicator:
       - ܻ
௜௧ = ܜܫܮ
௜௧ (1 if t > year of liberalization, 0 otherwise) plus country and time fixed effects (equation 5).
- Validation:
  - SWWW dates align with sharp changes in trade policy and increases in actual trade flows.
  - Fraser Institute’s Freedom to Trade Internationally Index (FTI) and its tariffs subcomponent show marked increases in liberalizing countries with a noticeable break around the SWWW date.
  - The typical liberalizing country started from a relatively low level of the FTI but reached or surpassed the world average within 10 years of liberalization (noted to be pronounced for LAC).
- Caveats:
  - World average is an imperfect control group because other countries may liberalize within the event window.
  - Even absent effects on actual openness, liberalization could affect outcomes via pro-competitive effects or technological transfers.

### Trade and Economic Performance: Main Findings
- Trade openness is beneficial for economic growth:
  - Benchmark specification (column 1, Table 1) finds trade openness has a positive and statistically significant impact on average per capita economic growth.
  - Evidence of conditional convergence: countries with lower initial real GDP per capita have higher average growth rates.
  - Labor force education coefficient is positive but not statistically significant.
  - Public infrastructure coefficient is positive and statistically significant.
  - Terms of trade have an adverse effect on real per capita growth.
  - Foreign direct investment has a positive but not statistically significant impact.
- Tests and instrument validity:
  - Second-order serial correlation tests indicate no serial correlation in the error term.
  - Hansen test of over-identifying restrictions on the full set of instruments confirms validity of the instrument set for the benchmark specification.

### Heterogeneity: Development Level and Region
- Interaction with advanced economy status:
  - Including an advanced economy dummy and interacting it with trade openness (column 2, Table 1) suggests the impact of trade openness on economic growth is stronger for advanced countries relative to countries at earlier stages of development.
  - Note: Hansen test results are indicative of weak instruments in this regression; conclusions should be interpreted with caution. Fixed effects regressions using lags yield consistent results.
- Regional variation:
  - LAC interaction (column 3, Table 1) suggests a stronger economic impact of trade for Latin America and the Caribbean relative to other regions.
  - A converse (not reported) result for an Asian dummy suggests a lower impact for Asian economies, lending support to the hypothesis that characteristics of countries’ trade connections matter.

### Key Regression Results (Table 1, system GMM; time fixed effects; robust SEs)
- Coefficients and statistics reported (columns 1–4):
  - Initial GDP per capita: -3.816***, -8.825***, -3.713***, -4.187*** (standard errors: (0.759), (2.961), (0.718), (0.907))
  - Labor force education: 0.773, -2.484, 0.478, -0.374 (standard errors: (0.560), (1.837), (0.526), (2.133))
  - Terms of trade: -1.741, 2.591, -1.495, -1.604 (standard errors: (1.833), (5.065), (1.688), (1.666))
  - Public Infrastructure: 1.674***, 5.094***, 1.895***, 0.643 (standard errors: (0.407), (1.813), (0.394), (1.019))
  - Trade openness: 2.045***, -0.0408, 1.658***, 0.426 (standard errors: (0.574), (1.494), (0.574), (1.298))
  - FDI (percent of GDP): 0.274, 0.430, 0.362*, 0.399 (standard errors: (0.238), (0.533), (0.217), (0.266))
  - Advanced economy (dummy): -38.05* (standard error: (22.07))
  - Advanced economy (dummy)*Trade Openness: 10.97** (standard error: (5.551))
  - Latin America and the Caribbean (dummy): -2.196*** (standard error: (0.814))
  - Latin America and the Caribbean (dummy)*Trade Openness: 6.723** (standard error: (2.968))
  - Labor force education*Trade Openness: 0.381 (standard error: (0.293))
  - Public Infrastructure*Trade Openness: 0.194 (standard error: (0.578))
- Sample and test statistics:
  - Observations: 1,041 in each column
  - Number of ISO: 135
  - AR(1): 0.0190, 0.0299, 0.0190, 0.0192
  - AR(2): 0.244, 0.309, 0.254, 0.241
  - Hansen: 0.5110, 0.006940, 0.9010, 0.577

### Trade Connectivity and Network Measures
- Rationale: Network indicators capture granular aspects of integration beyond trade openness, including number/diversity of trading partners (total degree, out-degree, in-degree) and centrality in the world trade network.
- Data limitations:
  - Available bilateral trade data include only trade in goods; services are not captured and network indicators may understate integration for service-intensive countries.
  - Network indicators computed from IMF Direction of Trade Statistics (DOTS) (goods only).
- LAC findings from network indicators:
  - LAC appears more integrated in the world trade network based on bilateral goods data than suggested by trade openness measures.
  - LAC maintains a diversified set of trading partners; larger Latin American economies occupy central roles in the world trade network.
  - South America countries are most integrated in LAC in terms of diversity of trading partners and centrality.
- Concept notes:
  - Total degree = number of a country’s total trading partners (exports + imports).
  - Out-degree = number of export partners; in-degree = number of import partners.
  - Eigenvector centrality (eigencentrality): standardized relative to the most central country and demeaned; centrality relates a country’s importance to both its own connections and the importance of those connections.
- Policy-relevant conclusion:
  - More expansive and diverse trading networks can enhance economic activity.

*Italic: Source: wp1746 - Section 2 (PDF) — IMF staff content provided in the input.*

### Section 3

### wp1746 - Section 3

### Trade Connectivity and Economic Growth: Key Findings
- Network measures (standardized relative to the most connected country in the world trade network and demeaned) matter for growth in addition to trade openness.
- Expanding a country’s trade connections yields economic benefits primarily through increasing export markets.
- Moving from the 25th percentile to the 75th percentile of total trade partners could improve average growth outcomes by about 3 percentage points over a five-year period; this corresponds to an increase of 142 trading partners.
- Increasing a country’s export markets (out degree) by 85 partners to move from the 25th to the 75th percentile would enhance average growth by 1.8 percentage points over a 5-year period.
- The number of export markets (Out Degree) is significant and positive for growth, while the number of import source countries (In Degree) is insignificant for growth.
- Overall centrality in the world trade network (Eigencentrality) is positively associated with growth; centrality among importers or exporters separately is less important than overall centrality.

### Empirical Results (selected coefficients from Table 2; system GMM growth regressions)
- Initial GDP per capita: -3.754***; -3.491***; -3.637***; -3.469*** (columns (1)–(4))
- Labor force education: -0.211; -0.336; -0.124; 0.102
- Terms of trade: -11.24***; -11.44***; -11.58***; -8.707*** 
- Public Infrastructure: 1.651***; 1.577***; 1.511***; 1.775***
- Trade openness: 1.255*; 1.028*; 1.333**; 1.336**
- FDI (percent of GDP): 0.636**; 0.672***; 0.543**; 0.638***
- Number of trading partners (Total degree): 7.695*** (column (1))
- Number of countries import from (In Degree): 3.992 (column (2))
- Number of countries export to (Out Degree): 3.846* (column (2))
- Centrality in the world trade network (Eigencentrality): 9.787*** (column (3))
- Centrality among exporters (Out eigencentrality): -5.719 (column (4))
- Centrality among importers (In eigencentrality): 10.50 (column (4))
- Observations: 829 (all columns)
- Number of ISO: 131 (all columns)
- AR(1): 0.0595; 0.0615; 0.0561; 0.0645
- AR(2): 0.390; 0.392; 0.395; 0.409
- Hansen: 0.199; 0.399; 0.171; 0.544
- Note: *** p<0.01, ** p<0.05, * p<0.1.

### Latin America and the Caribbean (LAC): Connectivity Patterns
- About 2/3 of LAC countries fall below the world average in total trading partners; much of this under-connectivity is concentrated in Caribbean countries.
- Small island economies (Eastern Caribbean Currency Union six independent members combined population around 600 thousand) are the least integrated in terms of number of trading partners; increasing trade connections may be infeasible for some small island economies.
- Argentina, Brazil, Colombia, Chile, and Peru expanded the number of their trading partners and are among the most diversified countries globally in total number of partners and export partners.
- Mexico and Panama are relatively less connected and may have potential to further diversify:
  - Mexico’s limited diversification linked to dominance of the U.S. market.
  - Panama’s limited connectivity may reflect importance of trade in services.
- Caribbean remains on the periphery of world trade network; prospects to become more central are likely limited given small size of many economies.

### Global Value Chains (GVCs): Concepts and Growth Effects
- GVC participation measured as the sum of foreign value added in gross exports and indirect value added (value of exported goods used as imported inputs by other countries to produce their exports), based on the Eora MRIO and Koopman et al. decomposition.
- Countries with higher foreign value-added are more downstream (backward-oriented); those with higher indirect value added are more upstream (forward-oriented).
- Increasing a country’s participation in GVCs from the 25th to the 75th percentile could enhance its 5-year average per capita growth by 1.8 percentage points.
- Upstream participation (indirect value added) has the more important effect on growth; downstream participation (foreign value added) has a positive but not statistically significant effect.
- The length of the GVC (captured by domestic value-added in intermediate goods production re-exported to third countries) also appears to matter for economic growth.

### Empirical Results on GVC Participation (selected coefficients from Table 3; system GMM growth regressions)
- Initial GDP per capita: -4.747***; -4.494***; -4.610*** (columns (1)–(3))
- Labor force education: -1.495; -1.447*; -1.620*
- Terms of trade: -8.084; -6.041; -7.280
- Public Infrastructure: 2.508***; 2.456***; 2.484***
- Trade openness: 2.920***; 3.144***; 3.350***
- FDI (percent of GDP): 0.463; 0.606; 0.675
- Global Value Chain Participation 2/: 0.0909** (column (1))
- FVA/Gross Exports: 0.0304; 0.0657 (column (2) entries)
- Indirect Value Added 3/: 0.0889* (column (2))
- Domestic Value Added in Intermediate Goods Exports Re-exported to Third Countries/Gross Exports: 0.485*** (column (3))
- Observations: 652 (all columns)
- Number of ISO: 135 (all columns)
- AR(1): 0.0197; 0.0169; 0.0178
- AR(2): 0.582; 0.611; 0.609
- Hansen: 0.0666; 0.0502; 0.0302
- Note: *** p<0.01, ** p<0.05, * p<0.1.
- Caution: Hansen test results indicative of weak instruments for regressions that include different measures of participation in GVCs; interpret results with caution.

### LAC Participation in GVCs: Patterns and Potential
- LAC’s participation in GVCs remains limited relative to other regions, particularly Asia.
- No clear regional trading hub in LAC comparable to China in Asia or Germany in Europe; the U.S. is a more pronounced hub for regional trade.
- Heterogeneity within LAC:
  - Mexico and Central America engage in production networks with North American firms and tend to be downstream/final-stage participants.
  - South American countries tend to be upstream/earlier-stage participants due to natural resource endowments.
  - LAC countries tend to participate in relatively shorter GVCs.
- Bringing LAC’s participation in GVCs in line with Asia could increase the region’s real per capita growth by about 1 percentage point.

### Policy Recommendations to Enhance Trade Connectivity and GVC Participation
- Remove trade barriers, with emphasis on the import of intermediate goods, because trade barriers are compounded across multiple border crossings in GVCs and act as a tax on exports.
- Address infrastructure deficiencies to strengthen participation in GVCs.
- Foster human capital formation to enable firms and workers to capture technological spillovers.
- Support research and development and improve institutions to facilitate a shift upstream in GVC participation and capture a higher share of value-added.
- Pursue targeted measures to broaden export markets (out degree) rather than focusing solely on import source diversification.

*Source: wp1746 - Section 3*

### Section 4

### Section 4

### Export markets and economic growth
- Trade connections with regional partners are associated with larger growth effects than inter-regional trade (Regional partners coefficient in Table 4: 1.019*** (0.369)).
- LAC regional export integration:
  - LAC: about a quarter of total exports destined to markets within the region; many LAC countries have regional export shares well below the 50th percentile of the world distribution.
  - Increasing intra-regional trade in LAC from about 15 percent of the region’s exports to 55 percent could enhance regional per capita growth by 1.3 percentage points.
  - LAC’s lower regional integration partly reflects its natural resource endowments and commodity-concentrated production structure.
- Historical and recent export market shifts:
  - Share of emerging markets and developing countries in LAC countries’ total exports increased about 50 percent, from 33 percent on average over 1995-1999 to about 50 percent on average over 2010-2015.
  - The expansion of trade with emerging markets and developing countries, driven in large part by China’s entry into the WTO and emerging Asia, may have lowered the overall impact of trade on LAC growth relative to historical ties with advanced economies.
- Centrality of trading partners:
  - Overall trade with the most central countries in the world trade network does not significantly matter for growth (Most central countries coefficient in Table 4: -0.251 (0.350)).
  - Trade with the most central importers is associated with stronger growth outcomes (Most central importers coefficient in Table 4: 0.549* (0.293)); increasing the share of trade with the most central importers from the 25th percentile to the 75th percentile would increase average per capita growth by 0.8 percentage points.
  - Concentration of export markets (concentration with 3 largest partners or export market diversification) does not show additional growth benefits (Concentration of trade with 3 largest trading partners: -1.492 (1.379); Export market diversification (HH index): -0.912 (0.871)).
- Formal trade agreements:
  - Number of trade agreements has a positive and significant association with growth (Number of trading partners with a trade agreement coefficient in Table 4: 2.296*** (0.544)).
  - Increasing the number of trade agreements from the 25th percentile to the 75th percentile is associated with a 1.5 percentage point increase in average per capita growth.
  - Evidence cited: Hannan (2016) finds trade agreements can increase exports by 80 percentage points over ten years on average for treated pairs.
  - Most LAC countries are above the 50th percentile in the global distribution of the number of trade agreements; Colombia and Mexico rank among the most integrated by this measure.

Policy-relevant implications from export-market findings:
- Deepening intra-regional trade links could yield sizeable per capita growth gains for LAC, but commodity specialization and production structure may limit possibilities.
- Strengthening ties with advanced economies and with the most central importers can enhance growth benefits from trade.
- Expanding and deepening formal trade agreements (including broader coverage beyond tariffs) can increase growth, especially for emerging markets.

### Export product characteristics and economic growth
- Export quality:
  - Higher export quality is linked to stronger growth effects of exports (Table 5, column 1).
  - Mexico, Central America, and the Caribbean tend to fare better on export quality; larger Latin American economies remain below the 50th percentile of the world distribution by this measure.
  - A stronger focus on export quality could strengthen the region’s gains from trade.
- Export diversification and concentration:
  - LAC has revealed comparative advantage in mineral fuels and non-fuel primary commodities and lags in skill- and technology-intensive manufactured products.
  - Some commodity exporters (e.g., Trinidad and Tobago and Venezuela) display relatively elevated export concentration; the Caribbean also shows high concentration in export product portfolios.
  - Increasing diversification of LAC’s export product basket by 25 percent has the potential to increase the region’s real per capita growth by about 1 percentage point.
  - However, diversification potential may be limited by comparative advantage in commodity exports due to natural resource endowments.
- Economic complexity:
  - Economic complexity has a positive, but not significant, effect on economic growth in the regressions reported.
- Measurement notes and data:
  - Export diversification is measured using the IMF Export Diversification Index (higher values indicate lower diversification / higher concentration).
  - Export quality is measured by the IMF Export Quality Index (higher values indicate higher export quality).
  - Economic complexity is measured using the Observatory of Economic Complexity.

Policy-relevant implications from export-product findings:
- Policies to improve export quality and to diversify the export product base can materially increase per capita growth in LAC.
- Constraints from natural-resource-based comparative advantage mean diversification and quality upgrading should be realistic and targeted, focusing on feasible value-added transitions.

*Source: wp1746 - Section 4*

### Section 5

### wp1746 - Section 5

### Event studies: growth, volatility and structural change
- Growth and investment tend to pick up after liberalization, supported by foreign capital inflows.
- The few years immediately preceding liberalization tend to be low-growth and high-volatility years; reforms are often preceded by downturns or crises.
- Growth and investment increase immediately after liberalization and the effects do not die out after a few years.
- Major trade reform episodes are associated with rising foreign direct investment (FDI), potentially related to accompanying policy changes that liberalize financial flows.
- A significant acceleration in FDI inflows may have contributed to growth through increases in the capital stock and technology transfer.
- The average liberalizing country also experienced a drop in economic volatility in the aftermath of liberalization.

Key reported panel-regression coefficients (selected, as reported in text and tables):
- Initial GDP per capita: -3.822***, -2.759***, -4.047***, -3.740***
- Public Infrastructure: 1.417***, 1.335***, 1.467**, 1.552**
- Trade openness: 1.860**, 0.0791, 1.705**, 1.223
- FDI (percent of GDP): 0.229, 0.507***, 0.294, 0.327
- Observations: 989, 784, 995, 969 (varies by specification)
- Number of ISO: 133, 110, 129, 129 (varies by specification)
- AR(1): 0.0271, 0.062, 0.0219, 0.0245
- AR(2): 0.386, 0.353, 0.238, 0.240
- Hansen: 0.233, 0.648, 0.219, 0.546

Sources and notes for event-study charts: WDI, Wacziarg and Welch (2008) and IMF staff calculations; cross-country averages with varying sample sizes by year.

### Trade liberalization and economic performance
- Aggregate impacts:
  - Growth tends to rise after trade is liberalized.
  - Investment strengthens after liberalization.
  - FDI inflows increase following trade liberalizations.
  - Economic volatility tends to decline following trade liberalizations.
- Structural effects:
  - Trade opening induces sectoral reallocations consistent with comparative advantage and access to new technologies.
  - Manufacturing export composition shifted toward goods of medium skill-intensity with a declining share of high-skill goods.
  - Production structure changes: decline in the share of agriculture in value added and an increase in the share of the tertiary (services) sector, moving the economic structure closer to the world average.
  - Such structural changes can produce winners and losers and may impose significant adjustment costs in segments of society (e.g., rural areas and some manufacturing industries).

Selected panel-regression evidence (Table 6 summary):
- Liberalization indicator effects (selected dependent variables):
  - Per capita growth: 2.663***
  - Investment rate: 2.307**
  - FDI inflow: -0.311 (not significant)
  - Volatility: -0.503 (not significant)
  - Agriculture share: -1.956
  - Services share: 3.432*
  - High-skill manufacturing share: -0.122***
  - Medium-skill manufacturing share: 0.082***

Sample and regression diagnostics (Table 6, selected):
- Number of observations (varies by dependent variable): 561, 1517, 857, 1135, 424, 470, 246, 895, 516, 55, 16 (as reported across table entries)
- Number of countries: 138, 137, 138, 138, 134, 134, 140, 140
- Adjusted R-squared (within): 0.087, 0.039, 0.047, 0.051, 0.316, 0.286, 0.047, 0.107

### Export composition and production structure (graphical evidence)
- Manufacturing exports: share of high-technology goods declines after liberalization; share of medium-technology goods rises.
- Value added composition: share of agriculture in value added declines faster after liberalization; share of services rises.
- The regression evidence largely confirms the graphical patterns: coefficients consistent with shift away from agriculture and toward services and medium-technology manufacturing (with exception that agriculture value added share coefficients were not significant in some regressions).

Data sources for composition and structure analysis: COMTRADE, UNCTAD, WDI, Wacziarg and Welch (2008) and IMF staff estimates.

### Trade and inequality: trends and regional patterns
- Time variation:
  - Over the past three decades, income inequality has shown substantial time variation while poverty levels generally declined.
  - On average, income inequality trended upward in advanced economies from the early 1990s.
  - Emerging and developing countries experienced substantial deterioration of inequality measures in the 1990s with some offsetting improvements more recently.
- Relationship between trade liberalization and inequality:
  - The sharp increase in inequality in emerging regions corresponds to a period of trade liberalization by many of these countries, but causal attribution is not established.
  - Poverty and inequality are distinct: falling poverty can coincide with rising inequality.
- Role of redistributive policies:
  - Redistributive policies substantially affect cross-sectional variation in income inequality.
  - LAC and Sub-Saharan Africa are the most unequal regions in net income terms.
  - Advanced economies and European developing countries use redistributive policies extensively to reduce income disparities from market incomes.
  - Governments in LAC appear less effective at using taxes and transfers to reduce high income inequality; policy recommendations include overhauling regressive tax systems, improving tax administration efficiency, and making social programs more targeted (Arnson, Bergman, and Fairfield, 2012).

Regional Gini observations (source: SWIID):
- Market-income median inequality: LAC below advanced economies and European emerging markets.
- After taxes and transfers (net Gini): LAC has the most unequal income distribution globally.

### Inequality regressions and event studies on liberalization
- Panel fixed effects regressions (five-year panels over 1980–2013) indicate:
  - Trade openness coefficient on Gini: negative but not statistically significant in most specifications.
  - Financial deepening is associated with rising inequality: Financial Deepening 0.0342*** (Market Gini) and 0.0215** (Net Gini) in one reported specification.
  - Financial openness has a positive but not significant effect on inequality in reported specifications.
  - Education and government spending coefficients reported but generally not significant in the presented regressions.
- Heterogeneous effects by development level:
  - Some evidence trade openness may decrease inequality for emerging market economies and increase it (not statistically significant) for advanced economies.
  - The negative effect of trade openness for emerging markets is significant only when market inequality is considered, implying a role for redistributive policy.
  - No distinct distributional effect detected for LAC relative to other regions in these regressions.
- Event studies on inequality around trade liberalization:
  - Newly liberalizing countries did not experience a larger increase in income inequality than the world average; they experienced a smaller increase in their Gini coefficient than closed economies.
  - For Latin America specifically, the rise in inequality after LAC trade reforms was much smaller than in the control group.
  - Simple panel regressions: coefficient on the liberalization indicator is insignificant for both market and net Gini (Table 8).
- Selected regression coefficients related to inequality (Table 7 and related specifications, selected):
  - Trade Openness: -0.0205, -0.00851, -0.0413*, -0.0216, -0.0220, -0.00796 (various specifications and Gini measures)
  - Financial Openness: 0.000985, 0.000638, 0.000456, 0.000305, 0.001010, 0.000629
  - Financial Deepening: 0.0342***, 0.0215**, 0.0349***, 0.0219**, 0.0343***, 0.0215**
  - Constant terms: 48.74***, 39.67***, 48.12***, 39.28***, 48.80***, 39.65***
  - Observations in these regressions: 562 (for each specification shown)
  - Number of ISO countries: 118 (for each specification shown)
  - R-squared: 0.133, 0.120, 0.143, 0.126, 0.133, 0.120 (various columns)

Event-study stylized comparisons:
- Trade liberalization and income inequality (change in Gini coefficient 10 years before and after trade liberalization):
  - Full sample (74 episodes): Newly liberalized countries’ changes in Market Gini and Net Gini followed global trends and were lower than closed economies.
  - LAC sample (21 episodes): The average liberalizing country in LAC fared better than the rest of the world in terms of changes in Gini.

Skill-premium evidence for LAC (SEDLAC, five countries: Argentina, Brazil, Panama, Uruguay, Venezuela):
- No clear trend in skill premium around trade liberalization episodes.
- Skill premium (high/low and medium/low) varied considerably over time and showed no noticeable break around liberalization episodes; almost none of the skill premium measures seem to increase after trade liberalization.

Selected regression table on liberalization and Gini (Table 8, summary):
- Dependent variables: Market Gini, Net Gini
- Liberalization coefficient examples: -0.447, 0.06 (insignificant in reported specifications)
- Number of observations: 3557 (Market Gini), 3557 (Net Gini)
- Number of countries: 138
- Adjusted R-squared (within): 0.198 (Market Gini), 0.14 (Net Gini)

### Conclusion and policy implications
- Consistent evidence across approaches:
  - Trade openness can promote economic growth without adversely affecting income inequality at the aggregate level.
  - Cross-country panel results: higher trade openness is associated with higher overall growth after controlling for other determinants.
  - Event studies: on average, growth and capital accumulation show favorable trends immediately after trade liberalization without increasing volatility.
  - Cross-country panel regressions: negligible or negative effect of trade integration on income inequality.
  - Structural change is associated with liberalization, but average income inequality in liberalizing and non-liberalizing countries followed virtually identical paths.
- Opportunity for Latin America and the Caribbean (LAC):
  - LAC can leverage trade to promote economic growth without increasing income inequality.
  - The region’s integration remains weak relative to other emerging markets; opportunities to enhance inter- and intra-regional trade could be explored to promote economic growth.

*Source: wp1746 - Section 5 (IMF staff analysis and underlying datasets cited within the section).*

### Section 6

### wp1746 - Section 6

### Trade integration and growth: qualitative findings
- With rising protectionist sentiment in advanced economies, increased regional integration may offer LAC the greatest potential to enhance the growth dividends of trade.
- The region could focus on promoting trade more generally rather than on specific aspects of trade per se, based on trade characteristics shown to promote economic growth.
- Continuing regional efforts to strengthen infrastructure and human capital is important to support a broader growth strategy and to enhance trade integration, including participation in global value chains.
- Policies to support export quality and promote export diversification can facilitate further gains from trade.

### Inclusion, inequality, and adjustment
- Trade can promote economic growth without adversely affecting overall income inequality, according to the analysis.
- Trade integration may lead to adjustment costs in particular segments of society; targeted social safety nets are warranted to support the adjustment process.
- The gap between the market and net Gini coefficients in LAC relative to other regions suggests that LAC has fallen far behind in the equity of its existing redistributive policies.

### Empirical regression highlights (Table A1: International Trade Connectivity and Economic Growth)
- Regression framework: regressions of growth in real GDP per capita on indicators of trade connectivity using system GMM. All regressions include time fixed effects. Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1.
- Sample size and panels:
  - Observations: 829
  - Number of ISO: 131
- Diagnostic tests (reported across specifications):
  - AR(1): values include 0.0595, 0.0615, 0.0626, 0.0593, 0.0595, 0.0587, 0.0620, 0.0561, 0.0645, 0.0175, 0.0292, 0.0317, 0.0331, 0.0329
  - AR(2): values include 0.390, 0.392, 0.411, 0.407, 0.406, 0.402, 0.400, 0.395, 0.409, 0.354, 0.326, 0.364, 0.322, 0.359
  - Hansen: values include 0.199, 0.399, 0.557, 0.217, 0.231, 0.182, 0.167, 0.171, 0.544, 0.768, 0.680, 0.818, 0.618, 0.708

- Selected coefficient estimates (coefficient estimate (standard error)):
  - Initial GDP per capita:
    - Ranges across specifications: -3.754*** (0.626); -3.491*** (0.579); -3.362*** (0.561); -3.733*** (0.654); -3.779*** (0.636); -3.706*** (0.615); -3.895*** (0.640); -3.637*** (0.603); -3.469*** (0.588); -2.985*** (0.609); -2.914*** (0.549); -3.145*** (0.574); -3.348*** (0.633); -3.184*** (0.536)
  - Labor force education:
    - Examples: -0.211 (0.562); -0.336 (0.527); -0.00321 (0.600); -0.104 (0.593); 0.102 (0.598); 0.221 (0.503); 0.225 (0.487); 0.146 (0.552); 0.0502 (0.546); 0.223 (0.559)
  - Terms of trade:
    - Examples: -11.24*** (3.006); -11.44*** (3.042); -7.934** (3.132); -8.192** (3.281); -9.247*** (3.201); -10.75*** (3.046); -10.65*** (3.106); -11.58*** (3.149); -8.707*** (3.052); -3.147 (2.418); -4.789* (2.557); -3.715* (2.039); -6.124** (2.543); -3.910** (1.814)
  - Public Infrastructure:
    - Examples: 1.651*** (0.404); 1.577*** (0.371); 1.788*** (0.359); 1.894*** (0.384); 1.813*** (0.389); 1.658*** (0.401); 1.768*** (0.390); 1.511*** (0.416); 1.775*** (0.369); 1.479*** (0.382); 1.036*** (0.388); 1.596*** (0.343); 1.364*** (0.370); 1.600*** (0.337)
  - Trade openness:
    - Examples: 1.255* (0.654); 1.028* (0.608); 1.185* (0.617); 1.424** (0.621); 1.457** (0.601); 1.328** (0.621); 1.461** (0.628); 1.333** (0.660); 1.336** (0.663); 1.561** (0.664); 1.059* (0.602); 1.147* (0.608); 1.307** (0.648); 1.244** (0.611)
  - FDI (percent of GDP):
    - Examples: 0.636** (0.261); 0.672*** (0.243); 0.647*** (0.251); 0.745*** (0.253); 0.713*** (0.257); 0.649** (0.257); 0.677*** (0.262); 0.543** (0.269); 0.638*** (0.231); 0.403* (0.239); 0.456* (0.238); 0.507** (0.230); 0.423* (0.254); 0.489** (0.228)

- Network connectivity measures (examples where reported):
  - Number of countries import from (In Degree): 3.992 (3.086)
  - Number of countries export to (Out Degree): 3.846* (2.232)
  - Number of trading partners (Total degree): 7.695*** (2.294)
  - Betweeness centrality: 2.525** (1.224)
  - Out betweeness centrality: 2.936** (1.149)
  - Closeness: 11.79*** (3.714)
  - Out closeness: 10.14*** (3.285)
  - Eigencentrality: 9.787*** (2.500)
  - Out eigencentrality: -5.719 (4.529)
  - In eigencentrality: 10.50 (6.676)
  - Power centrality: -7.292** (3.488)
  - Authority score: 9.857*** (2.889)
  - Authority score weighted: 6.193 (4.278)
  - Hub score: 6.567*** (1.921)
  - Hub score weighted: 4.197* (2.255)

### Policy implications and recommendations
- Prioritize regional integration as a pathway to enhance growth dividends of trade in an environment of rising global protectionism.
- Strengthen infrastructure and human capital as part of a broader growth strategy to boost trade integration and participation in global value chains.
- Promote export quality improvement and export diversification to realize additional trade gains.
- Implement targeted social safety nets to assist segments of society facing adjustment costs from trade integration.
- Address the equity gap in redistributive policies given the disparity between market and net Gini coefficients in LAC.

*Source: wp1746 - Section 6*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2017/wp1746.pdf_