## _wp1288

## Source details

**Canonical URL:** [_wp1288](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1288.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1288.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1288.pdf.json)

---

### Introduction and research question
- Objective: quantify the spillover effect of movements in China’s exchange rate on exports of other developing countries in third-country markets.
- Key conceptual point: analysis focuses on how a depreciation/appreciation of the renminbi affects exports of competitor countries in third markets (the “beggar-thy-neighbor” spillover).

### Data and empirical identification
- Period: 2000-2008.
- Trade data:
  - Disaggregated trade data at the 6-digit level covering roughly 6,000 non-oil HS 6-digit lines covering 900 HS 4-digit products.
  - Coverage: 124 developing country exporters and 57 large importers (57 major importing countries covering over 95 percent of total exports of developing countries).
  - Developing-country definition: World Bank low and middle income (2010 GNI per capita of $12,275 or less).
- Exchange rates:
  - Source: IMF IFS.
  - Core bilateral rate: nominal China-importer bilateral rate deflated by China’s CPI.
  - Alternative measures tested: nominal and real multilateral (IMF effective rate), internal real exchange rate from Penn World Tables (price level of GDP series).
- Indices of competition:
  - Value-Based Index (VBI) V_ijp: aggregates China’s share in importer j’s imports of each HS6 constituent g of HS4 product p, weighted by exporter i’s export composition across those HS6 lines (equation (11)).
  - Count-Based Index (CBI) C_ijp: fraction of HS6 lines within HS4 product p that both exporter i and China export to importer j (equation (12)).
  - Descriptive: both VBI and CBI rise over time; CBI for Asia exceeds 95 percent in 2008.

### Estimation strategy and identifying variation
- Core estimating equation (equation 13): regress log exports of HS 4-digit product p from exporter i to importer j on:
  - the log Chinese exchange rate vis-à-vis importer j (measured in renminbi per unit of j’s currency), and
  - the interaction of that exchange rate with an index of competition S_ijp (index does not have a time subscript).
- Fixed effects structure:
  - Importer-product-time fixed effects.
  - Exporter-product-time fixed effects.
  - Bilateral time-varying fixed effects and bilateral product-specific fixed effects.
- Identification intuition:
  - Compare exporters that differ in their degree of competition with China in the same third market; main variation comes from differences across exporters in S_ijp interacted with destination-specific Chinese exchange rate movements.
- Endogeneity mitigation:
  - Dependent variable is highly disaggregated exports (less likely to drive a macro exchange rate).
  - VBI and CBI constructed; indices computed for the initial period (year 2000) where relevant to mitigate endogeneity.

### Main findings and key statistics
- Sample size: N = 3,586,936 (core sample; nearly 3.6 million observations).
- Existence of a statistically and economically significant spillover:
  - A 10 percent appreciation of the renminbi increases a developing country’s exports at the product-level on average by about 1.5-2 percent.
  - For high indices of competition (e.g., 90th percentile), the increase could be as large as 2-3 percent using baseline β values and up to about 6 percent using larger β values (e.g., multilateral exchange rate measures).
- Baseline empirical magnitudes (median competition index, columns [4] and [8] of Table 1):
  - VBI: total spillover effect = -0.14.
  - CBI: total spillover effect = -0.20.
  - Interpretation: a 10 percent depreciation/appreciation of the renminbi is associated with a reduction/increase in developing-country exports at the product level to a third market of about 1.5-2 percent.
- Table 1 interaction coefficients (examples across columns): -0.178***; -0.227***; -0.128***; -0.352***; -0.250***; -0.234***; -0.158***; -0.222** with standard errors [ 0.002], [ 0.001], [ 0.001], [ 0.004], [ 0.001], [ 0.001], [ 0.001], [ 0.002].
- Representative theoretical parameter comparison:
  - Representative values: σ = 3, ω = 1, μ = 0.4, s = 0.4.
  - Using data averages: VBI average 0.4; CBI average 0.9.
  - Theoretical implied third-market spillover ≈ -0.32 (VBI) and -0.29 (CBI) versus empirical baseline -0.14 (VBI) and -0.20 (CBI).

### Robustness and sensitivity checks (summary of empirical evidence)
- Robustness to outliers, alternative clustering, long-difference (2000 vs. 2008), and alternative base years (2001, 2002): key coefficients remain negative and significant.
- Alternative competition measure: Finger-Kreinin export similarity index — results robust.
- Alternative exchange rate constructions:
  - Nominal bilateral (not deflated by China CPI) and real bilateral: coefficients remain correctly signed and significant though magnitudes decline relative to core.
  - Multilateral exchange rates (IMF effective nominal and real): coefficients remain significant and are substantially larger in magnitude.
  - Internal real exchange rate (Penn World Tables price level of GDP): interaction coefficient significant at 1 percent and larger in magnitude than core.
  - Selected coefficients in Table 3: -0.150***; -0.245***; 0.576***; 0.545***; 0.681***; -0.133***; -0.214***; 0.346***; 0.356***; 0.398** with standard errors [ 0.009], [ 0.009], [ 0.006], [ 0.006], [ 0.007], [ 0.004], [ 0.004], [ 0.003], [ 0.002], [ 0.003].
- Controlling for other countries’ exchange rate movements:
  - Identify top exporters whose currencies are highly correlated with renminbi (correlation thresholds 0.7, 0.4, 0.3) and interact those currencies with their competition indices.
  - Main China coefficient remains significant at 1 percent and slightly reduced in magnitude.
  - Spillover effects exist for other countries too but their competition indices are much lower than for China.
  - Table 4 interaction coefficients examples: -0.253***; -0.211***; -0.172***; -0.163***; -0.224***; -0.242*** with standard errors [0.005], [0.006], [0.008], [0.005], [0.006], [0.008].
- Geographic and aggregation robustness:
  - Results hold across exporter regions; coefficients larger for Asia than for Sub-Saharan Africa.
  - Table 5 examples (Value-based index): -0.467*** (Asia); -0.433*** (Europe); -0.297*** (LAC); -0.116*** (MENA+SSA).
  - Re-estimating at HS 2-digit level (N = 861,487): interaction term remains negative and significant.

### Product heterogeneity and mechanisms
- Greater spillover for homogeneous goods:
  - Classification: Rauch 1999 liberal classification concorded to HS6 and aggregated to HS4.
  - Interaction coefficients are larger (in absolute value) for homogeneous goods than for differentiated goods.
  - Magnitude differences: for the CBI and VBI, coefficients on homogeneous goods are about 20 percent and 40 percent greater, respectively, than for differentiated goods; differences statistically significant.
  - Table 7 examples: homogeneous/differentiated interaction coefficients -0.339***; -0.312***; -0.101***; -0.240***; -0.205***; -0.176***; interaction with Dummy for homogenous: -0.040***; -0.046***.
- Dampening with higher foreign-input content:
  - Hypothesis: higher imported intermediate content implies lower pass-through μ, damping spillovers.
  - Empirical finding: spillover effect is dampened for products with a high degree of foreign inputs.
  - Quantitatively: coefficient on interaction is about 13 percent greater (in absolute value) for the low-foreign-input sample for the value-based index and 10 percent greater for the count-based index; differences statistically significant.
  - Table 8 examples: index coefficients -0.329***; -0.285***; -0.283***; -0.236***; -0.191***; -0.170***; interaction with Dummy for high domestic value added: -0.125***; -0.100***.

### Discussion of magnitudes and economy-wide implication
- Definition: estimated spillover effect for exporter i, product p to importer j equals β multiplied by the competition index S_ijp.
- Baseline: a 10 percent renminbi move → about 1.5-2 percent change in developing-country product-level exports on average; 90th percentile competition → about 2-3 percent (baseline) and up to about 6 percent using larger β values.
- Back-of-envelope economy-wide implication:
  - Given an estimated average effect of about 1.5-2 percent per 10 percent renminbi move, and a 30 percent appreciation of China’s real exchange rate vis-à-vis the US dollar over 2000-2008, this implies a potential 4.5-6 percent increase in the typical developing country’s exports to the US associated with that appreciation — with larger effects for exporters in closer competition with China.

### Conclusion and policy implications
- Main conclusions:
  - This paper is the first attempt (to the authors' knowledge) to quantify the effect of exchange rate changes on the exports of competitor countries to third markets exploiting disaggregated trade data with a structural identification strategy.
  - China’s exchange rate changes can have significant and robust spillover effects on competitor countries’ exports to third markets.
- Policy implications:
  - If exchange rate movements stem from policy actions, the identified spillover effects imply that one country’s policies can have substantial export implications for other countries.
  - Findings have potential implications for developing countries and the multilateral system because of cross-border externalities from country-specific exchange rate policies.
- Limitations and open questions:
  - The analysis isolates the spillover channel but does not measure other channels that could benefit developing-country exports to China (e.g., increased demand for raw materials/intermediates to China, or China’s growth effects).
  - The net effect of China’s exchange rate on overall exports of other countries remains an open question.
  - Paper does not directly estimate effects of China’s exchange rate movements on China’s own exports; further research needed.

*Source: _wp1288 (Appendix Tables; Sections II–V; Conclusion; References) — content as provided.*

### Appendix Tables

### _wp1288 - Appendix Tables

### Introduction and research question
- Objective: quantify the spillover effect of movements in China’s exchange rate on exports of other developing countries in third-country markets.
- Motivation:
  - China is the world’s largest exporter of goods and a highly diversified exporter, likely to have quantitatively significant competitive effects on other countries.
  - China’s exchange rate policy has been a central international macroeconomic issue during the 2000s, with implications for demand and output in industrial countries.
- Key conceptual point: the analysis focuses on how a depreciation/appreciation of the renminbi affects exports of competitor countries in third markets (the “beggar-thy-neighbor” spillover).

### Data and empirical identification
- Data:
  - Disaggregated trade data at the 6-digit level.
  - Coverage: 124 developing country exporters and 57 large importers.
  - Period: 2000-2008.
- Identification strategy:
  - An analytical framework based on Feenstra, Obstfeld, and Russ (2011) motivates the approach.
  - Constructs indices of competition between China and other exporting countries at the exporter-importer-product level.
  - Core reasoning: the more a country competes with China in a third market, the more a given depreciation of the renminbi is likely to hurt its exports in that market.
- Econometric specification:
  - Employs a battery of fixed effects that control for importer-exporter-product, importer-exporter-time, exporter-product-time, and importer-product-time varying characteristics.
  - Fixed effects help address omitted variable bias common in aggregated trade–exchange rate estimations.
  - Reverse causality concerns are mitigated because disaggregated exports are unlikely to affect a macroeconomic variable like China’s exchange rate.

### Main findings and key statistics
- Existence of a statistically and economically significant spillover effect:
  - A 10 percent appreciation of the renminbi increases a developing country’s exports at the product-level on average by about 1.5-2 percent.
  - For high indices of competition, the increase could be as large as 6 percent.
- Interpretation:
  - An appreciation of the renminbi could provide a substantial boost to developing country exports going forward.
- Heterogeneity of effects:
  - The spillover effect is greater for homogeneous products with greater substitution possibilities than for differentiated products.
  - The spillover effect is attenuated for products that rely more on foreign inputs (and hence have a lower degree of Chinese domestic value added).

### Robustness and additional considerations
- Robustness checks and scope:
  - Estimates are robust to a variety of statistical tests.
  - Results hold for alternative measures of exchange rates.
  - Results hold for alternative disaggregation of trade data.
  - Results hold across exporting and importing regions.
  - Results are robust to incorporating competition from countries (other than China) whose currencies move with the renminbi.
- Causal interpretation strengthened by disaggregation and fixed effects structure; reverse causality less likely.

### Relation to existing literature and contribution
- Novelty: provides direct evidence on the effect of China’s exchange rate changes on exports of other countries in third markets — an area with limited prior evidence.
- Prior related studies:
  - Several studies examine China’s export performance and effects on other Asian countries but do not focus on exchange rates (Hanson and Robertson (2008), Eichengreen, Rhee and Tong (2004), Ahearne et. al. (2003)).
  - Other papers examine effects of exchange rate changes on variables other than trade (e.g., Eichengreen and Tong (2011) on stock market valuations).
- Contribution: combines highly disaggregated trade data with a structural identification strategy to quantify spillovers and link magnitudes to theoretical predictions.

### Organization of the paper (as provided)
- Analytical framework and further development presented in Section II.

*Source: _wp1288 - Appendix Tables (PDF chapter/section), content as provided.*

### Section  III  elaborates  the  estimation  strategy  and  Section  IV  describes  the  data.  The  results  are

### III. ESTIMATION STRATEGY

### Estimation approach and identifying variation
- Core estimating equation (equation 13) regresses log exports of HS 4-digit product p from exporter i to importer j on:
  - the log Chinese exchange rate vis-à-vis importer j (measured in renminbi per unit of j’s currency) and
  - the interaction of that exchange rate with an index of competition S_ijp (the index does not have a time subscript).
- Key intuition: Compare exporters that differ in their degree of competition with China in the same third market (e.g., Brazil vs. Malawi exports to the US). When the renminbi depreciates, exporters more exposed to China in that market lose relatively more market share.
- Fixed effects structure:
  - Importer-product-time fixed effects (captures importer-product-time shocks, e.g., fiscal support in an importing country).
  - Exporter-product-time fixed effects (captures exporter-product-time shocks, e.g., exporter productivity shocks).
  - Bilateral time-varying fixed effects and bilateral product-specific fixed effects capture bilateral factors (currency unions, exchange rate pegs, product-specific trade policy).
- Main remaining source of variation exploited: differences across exporters in the index of competition with China, interacted with destination-specific Chinese exchange rate movements.
- Endogeneity and reverse causality:
  - Dependent variable is highly disaggregated exports (less likely to drive a macro exchange rate).
  - Index constructions: count-based index (CBI) relies on extensive margin (less value-endogeneity); value-based index (VBI) uses values and may be more endogenous, so indices are computed for the initial period (year 2000) to mitigate endogeneity.

### Index of competition (theoretical motivation)
- Value-Based Index (VBI): A "value-based index of competition" (V_ijp) aggregates China’s share in importer j’s imports of each HS6 constituent g of HS4 product p, weighted by exporter i’s export composition across those HS6 lines. Appears in equation (11).
- Count-Based Index (CBI): A "count-based index" C_ijp equals the fraction of HS6 lines within HS4 product p that both exporter i and China export to importer j (an extensive-margin measure). Appears in equation (12).
- The spillover elasticity for exporter i, product p to importer j with respect to China’s exchange rate is proportional to:
  - the index of competition (VBI or CBI),
  - (σ_p − ω_p) where σ_p is the elasticity of substitution across imported varieties and ω_p is the elasticity between domestic and imported varieties,
  - and the pass-through μ_p from China’s domestic prices to importer-market prices.

### IV. DATA

- Sample period: 2000-2008.
- Trade data:
  - Source: UN Comtrade (imports reported by importing countries).
  - Coverage: roughly 6,000 non-oil HS 6-digit lines covering 900 HS 4-digit products.
  - Countries: 57 major importing countries (covering over 95 percent of total exports of developing countries) and 124 developing-country exporters potentially competing with China.
  - Developing-country definition: World Bank low and middle income (2010 GNI per capita of $12,275 or less).
- Data treatment:
  - Trade values reported in current US dollars and deflated by US CPI.
  - Acknowledged limitation: lack of bilateral price deflators; implicitly deflated via very general fixed effects (importer-product-time; importer-product-exporter; exporter-product-time).
- Exchange rates:
  - Source: IMF IFS.
  - Bilateral exchange rate used in core: nominal China-importer bilateral rate deflated by China’s CPI (motivated by the model where Chinese export price depends on China domestic price and bilateral exchange rate with pass-through).
  - Alternative exchange rate measures tested: nominal and real multilateral (IMF effective rate), and internal real exchange rate from Penn World Tables (price level of GDP series).
- Additional data notes:
  - Indices of competition (VBI and CBI) are constructed and averaged across exporters/products for descriptive analysis.
  - Figures (described): both VBI and CBI rise over time; CBI for Asia exceeds 95 percent in 2008.

### V. RESULTS — Main findings and robustness

### Baseline estimates
- Core sample size: nearly 3.6 million observations.
- Specifications:
  - Columns [1]-[4] use VBI; columns [5]-[8] use CBI.
  - Fixed effects progressively expanded; columns [4] and [8] represent the most comprehensive (core) specifications.
  - Standard errors clustered at importer-exporter-product level.
- Main empirical regularity:
  - The coefficient on the interaction between the Chinese exchange rate and the index of competition is consistently negative and significant at the 1 percent confidence level across specifications and both indices (VBI and CBI).
  - Interpretation: A depreciation of the renminbi vis-à-vis an importer’s currency is associated with a larger decline in a developing country’s exports of a product to that importer the more that developing country competes with China in that product-market.

### Robustness checks (summary)
- Outliers: Dropping top and bottom 1 percentile leaves key coefficient negative and significant with similar magnitude.
- Standard error clustering at higher aggregation levels (exporter-importer-year; exporter-product-year; importer-product-year) does not change statistical significance.
- Long-difference (2000 vs. 2008): results remain similar; interaction coefficient increases slightly.
- Index measured in alternative base years (2001, 2002): results robust.
- Alternative competition measure: Finger-Kreinin export similarity index — results robust.
- Alternative exchange rate constructions:
  - Nominal bilateral (not deflated by China CPI) and real bilateral: coefficients remain correctly signed and significant though magnitudes decline relative to core.
  - Multilateral exchange rates (IMF effective nominal and real): coefficients remain significant and are substantially larger in magnitude.
  - Internal real exchange rate (Penn World Tables price level of GDP): interaction coefficient significant at 1 percent and larger in magnitude than core.
- Controlling for other countries’ exchange rate movements:
  - Identify top exporters whose currencies are highly correlated with renminbi (correlation thresholds 0.7, 0.4, 0.3) and interact those currencies with their competition indices.
  - Main China coefficient remains significant at 1 percent and slightly reduced in magnitude.
  - Spillover effects exist for other countries too (negative coefficients), but indices of competition for other countries are much lower than for China, so overall impact is smaller.
- Geographic robustness:
  - Splitting the sample across exporter regions (Asia, sub-Saharan Africa, etc.): results hold across regions; coefficients larger for Asia than sub-Saharan Africa.
- Product aggregation robustness:
  - Re-estimating at HS 2-digit level (sample shrinks from over 3.6 million to about 860,000 observations): interaction term remains negative and significant.

### Spillover effect and product characteristics

- Theoretical determinants: Elasticity of substitution across imported varieties (σ_p) and exchange-rate pass-through (μ_p). Higher σ_p and higher μ_p imply larger spillover.
- Empirical tests:
  - Homogeneous vs. differentiated goods (Rauch 1999 classification; Rauch liberal classification used, concorded to HS6 and aggregated to HS4):
    - Interaction coefficients are larger (in absolute value) for homogeneous goods than for differentiated goods.
    - Magnitude differences: for the CBI and VBI, coefficients on homogeneous goods are about 20 percent and 40 percent greater, respectively, than for differentiated goods; differences statistically significant.
  - Dependence on foreign intermediate inputs (Koopman, Wang and Wei classification):
    - Hypothesis: Higher imported intermediate content implies lower pass-through μ, which should dampen spillover.
    - Empirical finding: Spillover effect is dampened for products with a high degree of foreign inputs.
    - Quantitatively: coefficient on interaction is about 13 percent greater (in absolute value) for the low-foreign-input sample for the value-based index and 10 percent greater for the count-based index; differences statistically significant.

### Discussion of magnitudes

- Definition: Estimated spillover effect for exporter i, product p to importer j equals β multiplied by the competition index S_ijp, where β is the estimated coefficient on the interaction.
- Baseline empirical magnitudes (median competition index):
  - Using baseline elasticity estimates (columns [4] and [8] of Table 1) and median index values:
    - VBI: total spillover effect = -0.14.
    - CBI: total spillover effect = -0.20.
  - Interpretation: A 10 percent depreciation/appreciation of the renminbi is associated with a reduction/increase in developing-country exports at the product level to a third market of about 1.5-2 percent.
- High-competition cases:
  - For countries in the 90th percentile of competition with China, baseline range increases to about 2-3 percent for a 10 percent renminbi movement.
  - Using larger β values (e.g., multilateral exchange rate measures), the spillover effect for 90th percentile competition could be as high as 6 percent for a 10 percent renminbi change.
- Theory vs. empirical:
  - Representative parameter values used for theoretical comparison: σ = 3, ω = 1, μ = 0.4, s = 0.4.
  - Using average competition indices from data: VBI average 0.4; CBI average 0.9.
  - Theoretical implied third-market spillover ≈ -0.32 (VBI) and -0.29 (CBI).
  - Empirical baseline estimates: -0.14 (VBI) and -0.20 (CBI).
  - Explanations for gap (VBI): measurement error attenuation bias; differences in relevant elasticities at HS4 aggregation versus micro estimates (Feenstra et al. use near HS10-level elasticities).
- Economy-wide implication (back-of-envelope):
  - Given an estimated average effect of about 1.5-2 percent per 10 percent renminbi move, and a 30 percent appreciation of China’s real exchange rate vis-à-vis the US dollar over 2000-2008, this implies a potential 4.5-6 percent increase in the typical developing country’s exports to the US associated with that appreciation — with larger effects for exporters in closer competition with China.

*Italic: Source — Sections II–V of the supplied IMF working paper content (estimation strategy, data, results, and magnitudes).*

### CONCLUSION

### _wp1288 - CONCLUSION

### Main findings
- This paper is the first attempt, to the authors' knowledge, to quantify the effect of exchange rate changes on the exports of competitor countries to third markets (the “spillover effect”) while exploiting rich variation from disaggregated trade data and doing so in a manner motivated by and consistent with theory.
- The case study of China finds that its exchange rate changes can have significant and robust spillover effects on competitor countries’ exports to third markets.
- The paper identifies precisely and robustly a specific mechanism of influence: Chinese exchange rate movements → spillover effect on exports of competitor countries to third markets.

### Policy implications
- If exchange rate movements (or the lack thereof) stem from policy actions, the identified spillover effects imply that one country’s policies can have substantial export implications for other countries.
- These findings could have important implications for developing countries and for the multilateral system, given potential cross-border externalities from country-specific exchange rate policies.

### Limitations and open questions
- The analysis isolates the spillover channel but does not measure other potentially beneficial channels affecting developing country exports to China:
  - A depreciation of the renminbi could increase developing country exports of raw materials and intermediate goods to China used in production of China’s exports to third countries.
  - A renminbi depreciation could boost China’s own growth, raising its demand for goods and services and thereby increasing developing country exports to China.
- The net effect of China’s exchange rate on overall exports of other countries remains an open question.
- The paper does not directly estimate effects of China’s exchange rate movements on China’s own exports; further research is needed to precisely identify these additional effects.

### Related literature (as cited)
- Goldberg and Knetter (1997) and Goldberg and Hellerstein (2008) provide evidence on pass-through and its decline over time. Xing (2010) estimates pass-through coefficients of 0.23 and 0.56 for the US and Japan respectively for Chinese exchange rates to import prices in US and Japan.23
- Broda and Weinstein (2006) argue that more disaggregated data can yield higher estimates of the elasticity of substitution.24

*Source: CONCLUSION, _wp1288*

### REFERENCES

### _wp1288 - REFERENCES

### References (selected as listed in source)
- Acemoglu, Daron and Simon Johnson, 2007. "Disease and Development: The Effect of Life Expectancy on Economic Growth," Journal of Political Economy, vol. 115(6), pages 925-985.
- Ahearne, Alan, John Fernald, Prakash Lougani, and John Schindler, 2003. “China and Emerging Asia: Comrades or Competitors?” International Finance Discussion Paper No. 789, (Washington, D.C.: Federal Reserve Board)
- Berman, Nicolas, Philippe Martin and Thierry Mayer, 2012. "How do different exporters react to exchange rate changes? Theory, empirics and aggregate implications," Quarterly Journal of Economics, 127(1): pp. 437-492.
- Broda, Christian and David E. Weinstein. 2006. "Globalization and the Gains from Variety," Quarterly Journal of Economics 121(2): 541-585.
- Campa, Jose Manuel, and Linda S. Goldberg, 2005. "Exchange Rate Pass-Through into Import Prices," The Review of Economics and Statistics, vol. 87(4), pages 679-690, December
- Das, S., Roberts, M., and J. Tybout, 2001. "Market Entry costs, Producer Heterogeneity, and Export Dynamics." NBER Working Paper No. 8629
- Dekle, Robert and Heajin H. Ryoo, 2007. “Exchange rate fluctuations, financing constraints, hedging, and exports: Evidence from firm level data.” Journal of International Financial Markets, Institutions and Money 17, 437–451
- Dooley, Michael P., David Folkerts-Landau, and Peter Garber, 2003. “An Essay on the Revived Bretton Woods System.” NBER Working Paper No. 9971
- Eichengreen, B., Y. Rhee and H. Tong (2004). “The Impact of China on the Exports of Other Asian Countries,” NBER Working Paper no.10768
- Eichengreen, Barry and Hui Tong, 2011. “The External Impact of China's Exchange Rate Policy: Evidence from Firm Level Data.” IMF Working Paper, 155
- Engel, Charles, 2009. “Exchange Rate Policies.” Federal Reserve Bank of Dallas Staff Paper.
- Feenstra, Robert C., Maurice Obstfeld, and Katheryn N. Russ, 2011. “In Search of the Armington Elasticity,” mimeo, UC Berkeley and UC Davis.
- Finger, J. Michael, and Mordechai E. Kreinin. 1979. "A Measure of 'Export Similarity' and Its Possible Uses," Economic Journal, 89, (December) pp. 905-912.
- Forbes, K., 2002. "How Do Large Depreciations Affect Firm Performance", NBER Working Paper No. 9095
- Goldberg, P and M. Knetter, 1997, “Goods Prices and Exchange Rates: What Have We Learned? Journal of Economic Literature, pp. 1243-72.
- Goldberg, P. and R. Helllerstein, 2008, “A Structural Approach to Explaining Incomplete Exchange Rate Pass-through and Pricing-To-Market,” American Economic Review, pp. 423-29.
- Goldstein, Morris and M. S. Khan, 1985. "Income and Price Effects in Foreign Trade", Handbook of International Economics, vol. 2. R.W. Jones and P.B. Kenen, eds. Amsterdam: North-Holland.
- Gopinath, Gita, Oleg Itskhoki, and Roberto Rigobon. 2011. “Currency Choice and Exchange Rate Pass-Through.” American Economic Review, vol. 100(1), pages 304-36
- Hooper, P., K. Johnson, and J. Marquez, 2000. "Trade Elasticities for G-7 Countries”, Princeton Studies in International Economics No. 87.
- Hanson, Gordon H., and Raymond Robertson, 2008. “China and the Manufacturing Exports of Other Developing Countries.” NBER Working Paper No. 14497.
- Johnson, Simon H., Jonathan Ostry, and Arvind Subramanian, 2010. "Prospects for Sustained Growth in Africa: Benchmarking the Constraints," IMF Staff Papers, vol. 57(1), pages 119-171, April
- Koopman, Robert, Zhi Wang, and Shang-Jin Wei, forthcoming, “Estimating domestic content in exports when processing trade is pervasive”, Journal of Development Economics.
- Rauch, J.E., 1999. “Networks versus Markets in International Trade.” Journal of International Economics 48, 7–35
- Robinson, Joan. 1947. Essays in the Theory of Employment, 2nd edition, Oxford: Basil Blackwell. First published in 1937.
- Rodrik, Dani, 2008. “The Real Exchange Rate and Economic Growth,” Brookings Papers on Economic Activity, 2, pp. 365–412.
- Rogoff, K., 1996. “The Purchasing Power Parity Puzzle,” Journal of Economic Literature 34, 647–668.
- Thursby J.G., and M.C. Thursby, 1987. “Bilateral Trade Flows, the Linder Hypothesis, and Exchange Risk,” Review of Economics and Statistics 69, 488-495
- Wei, Shang-Jin and Xiaobo Zhang, 2011, “The Competitive Savings Motive: Evidence from Rising Sex ratios and Savings Rates in China,” Journal of Political Economy, 119 (3), 511-564.
- Xing, Yuqing, 2010. “The Yuan's Exchange Rates and Pass-through Effects on the Prices of Japanese and the US Imports.” Comparative Economic Studies, 52, 531–548.

### Notes and Table Highlights (as presented in source)
- Notes on indices:
  - Value-based index: summation over all 6-digit products within the 4-digit category of: share of China in overall imports of a 6 digit product multiplied by the share of the 6-digit product in total 4-digit exports from the exporter to the importer.
  - Count-based index: at the 4-digit product level, equal to the share of 6-digit products within a 4-digit category that i exports to j, that China also exports to j.
- Table 1 (Exports from Developing Countries and Chinese Exchange Rates: Baseline Specification)
  - Index of competition with China*log(exchange rate of importer with respect to China): - 0.178***; - 0.227***; - 0.128***; - 0.352***; - 0.250***; - 0.234***; - 0.158***; - 0.222**
  - Standard errors shown in brackets: [ 0.002], [ 0.001], [ 0.001], [ 0.004], [ 0.001], [ 0.001], [ 0.001], [ 0.002]
  - N = 3,586,936
  - Fixed effects: exporter*importer*product (various combinations of exporter*importer*time, exporter*product*time, importer*product*time specified across columns)
  - Regression sample includes years from 2000-2008. Standard errors clustered at the importer*exporter*product level. ***, ** and * denote statistical significance at the 1, 5 and 10 percent respectively.
- Table 2 (Robustness)
  - Value-based index coefficient examples: -0.370***; -0.352***; - 0.352***; -0.352***; -0.416***; -0.326***; -0.311***; -0.385**
    - Standard errors: [0.005], [0.001], [0.003], [ 0.003], [0.038], [0.004], [0.004], [0.003]
  - Count-based index coefficient examples: -0.192***; -0.222***; -0.222***; -0.222***; -0.208***; -0.387***; -0.417***
    - Standard errors: [0.002], [0.001], [0.001], [ 0.001], [0.017], [0.002], [0.003]
  - N values: 3,479,214; 3,586,936; 3,586,936; 3,586,936; 788,775; 3,586,936; 3,586,936; 3,586,936
  - Robustness checks include dropping outliers, alternative clustering, long-difference (2000, 2008), alternative index years (2001, 2002), and Finger-Krenin index.
- Table 3 (Robustness to Alternative Exchange Rate Measures)
  - Selected coefficients: - 0.150***; - 0.245***; 0.576***; 0.545***; 0.681***; - 0.133***; - 0.214***; 0.346***; 0.356***; 0.398**
  - Standard errors: [ 0.009], [ 0.009], [ 0.006], [ 0.006], [ 0.007], [ 0.004], [ 0.004], [ 0.003], [ 0.002], [ 0.003]
  - N values: 3,586,936; 3,586,936; 3,602,228; 3,602,228; 3,602,228; 3,586,936; 3,586,936; 3,602,228; 3,602,228; 3,602,2
  - Notes: variety of nominal/real and effective exchange rate measures used (including IMF series and Penn World Tables price level p series). Increase in real and nominal effective exchange rates denotes an appreciation.
- Table 4 (Robustness to Omitted Variables)
  - Index of competition with China*log(exchange rate of importer with respect to China): - 0.253***; - 0.211***; - 0.172***; - 0.163***; - 0.224***; - 0.242***
  - Standard errors: [0.005], [0.006], [0.008], [0.005], [0.006], [0.008]
  - N: 3,141,707; 2,697,281; 2,248,111; 3,141,707; 2,697,281; 2,248,111
  - Controls include indices of competition with other countries (Malaysia, Poland, Mexico, Thailand, Brazil, India) depending on correlation thresholds.
- Table 5 (Robustness By Region of Exporter)
  - Value-based index examples: -0.467*** (Asia); -0.433*** (Europe); -0.297*** (LAC); -0.116*** (MENA+SSA)
  - Count-based index examples: -0.288*** (Asia); -0.258*** (Europe); -0.192*** (LAC); -0.139** (MENA+SSA)
  - N: 1,234,019; 997,174; 750,565; 436,403 (for both index types)
  - Regions defined by World Bank classification. LAC = Latin America and the Caribbean; MENA = Middle East and North Africa; SSA = Sub-Saharan Africa.
- Table 6 (Robustness to Degree of Product Aggregation, HS 2-Digit Level)
  - Index coefficients: -0.131***; -0.084***; - 0.009***; -0.306***; -0.293***; -0.206***; -0.113***; - 0.268**
  - N = 861,487 (across columns)
  - Notes: index defined over 2-digit category as summation over 6-digit products; sample years 2000-2008.
- Table 7 (Products Distinguished by Degree of Differentiation)
  - Homogeneous / Differentiated / Interaction coefficients (examples): -0.339***; -0.312***; -0.101***; -0.240***; -0.205***; -0.176***
  - Interaction with Dummy for homogenous: -0.040***; -0.046***
  - N values: 981,310; 2,679,680; 1,326,035; 981,310; 2,679,680; 1,326,035
  - Goods classified by Rauch's liberal classification at 6-digit level.
- Table 8 (Products Distinguished by Domestic Value Added)
  - Index coefficients examples: -0.329***; -0.285***; -0.283***; -0.236***; -0.191***; -0.170***
  - Interaction with Dummy for high domestic value added: -0.125***; -0.100***
  - N values: 1,511,450; 1,830,310; 3,341,760; 1,738,687; 1,848,249; 3,341,760
  - Regressions restricted to manufacturing. Domestic value added data from Koopman, Wang and Wei (forthcoming). High/low defined by median.
- Table 9 (Range of Estimated Spillover Effect of a 10 percent Depreciation of Chinese Exchange Rate)
  - Percentile of the index of competition (Value-based index / Count-based index)
    - 10- 0.010.00- 0.010.000.000.00
    - 50- 1.30-0.47- 2.52-1.99-1.39- 3.74
    - 90- 3.12-1.13- 6.03-2.22-1.55- 4.17
  - Notes: baseline, minimum and maximum coefficients correspond to specified columns across Tables 1, 2, 3, and 4 as indicated in source.
- Appendix: Steps in the Derivation of Equation (10)
  - Derivation starts from Equation (9) and applies calculus identities and substitution from Equation (8) to show algebraic steps resulting in Equation (10) in the paper. (Source reproduces the mathematical steps and identities applied.)
- Table A1. Summary Statistics (verbatim as in source)
  - Variable Observations Mean Standard Deviation Minimum Maximum
  - Nomi nal Ex ports ( '000 USD) 3,586,936 2009.79739531.45 00.001 1.590E+07
  - Log ( re al e x ports, de f l ate d by US CPI) 3,586,936 -1.4853.13 4 -11.61112.100
  - Index of competition with China (structural measure ) 3,586,936 0.4080.325 0.0001.000
  - Index of competition with China (count-based measure ) 3,586,936 0.8980.282 0.0001.000
  - Nominal exchange rate (renminbi / importer currency) 3,586,936 2.6143.332 0.00015.222
  - Log (renminbi/importer currency exchange rate, deflated by Chinese CPI) 3,586,936 -5.8922.426 -13.247- 2.632
- Table A2. List of countries (Exporting countries / Importing countries)
  - Full country lists for exporters and importers as provided in source (includes Afghanistan, Albania, American Samoa, Argentina, Armenia, Bangladesh, Belarus, etc., and importing countries including Macedonia, FYR; Madagascar; Algeria; Australia; Austria; Belarus; Belgium; Brazil; Canada; Chile; Colombia; Czech Republic; Denmark; Egypt, Arab Rep.; Finland; France; Germany; Greece; Hong Kong, China; Hungary; India; Indonesia; Ireland; Israel; Italy; Japan; Kazakhstan; Korea, Rep.; Malaysia; Mexico; Morocco; Netherlands; New Zealand; Nigeria; Norway; Pakistan; Philippines; Poland; Portugal; Qatar; Romania; Russian Federation; Saudi Arabia; Singapore; Slovak Republic; South Africa; Spain; Sweden; Switzerland; Taiwan, China; Thailand; Turkey; United Arab Emirates; United Kingdom; United States; Venezuela; Vietnam; and many others as listed.)

*Italic: Content unit _wp1288 - REFERENCES (source PDF filename: _wp1288 - REFERENCES).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1288.pdf_