## _wp06155 — 2. SensitivityAnalysis:TradeBalance

## Source details

**Canonical URL:** [_wp06155 — 2. SensitivityAnalysis:TradeBalance](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06155.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06155.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06155.pdf.json)

---

### Introduction and research question
- Literature reviewed: Harrigan (2000); Harrigan and Balaban (1999); Tokarick (2005); Feenstra and Hanson (2001).
- Central question: Do trade policies (tariff changes, FTAs) have negligible effects on domestic relative wages if they affect world prices? Net effect depends on whether trade policy affects domestic prices via world/regional prices.
- Paper advances:
  - Constructs a general equilibrium model for a “region” where member countries are large enough to affect regional world prices.
  - Uses 1995 international input-output data for 9 APEC member countries (plus associated regions), distinguishing tradables vs nontradables and final vs intermediate trade.
  - Key simulation: abolish observed 1995 within-region tariff rates (set to zero).

### Model and methodology
- Structure:
  - M countries, I industries; (M+1)th country = rest of the world.
  - Producer: X^n_j = f^n_j(x^1n_j,...,x^Mn_j, L^n_j, H^n_j; x̄^{M+1,n}_j, K̄^n_j) with x̄ and K̄ fixed.
  - Import prices include tariff and transport: p^m_i (1 + τ^mn_i) where (1 + τ^mn_i) = (1 + d^mn_i)(1 + t^mn_i).
  - Competitive equilibrium: prices (p^1,...,p^M, w_L, w_H) and allocations satisfy firms’ profit maximization, households’ utility maximization, and market clearing for goods and two labor types.
- Linearization and comparative statics:
  - Excess supply e and Jacobian Φ_p of excess supply w.r.t. prices and wages.
  - Small change in shift parameters μ satisfies Φ_p · Γ_μ + Φ_μ = 0 → Γ_μ = −Φ_p^{−1} Φ_μ.
  - Semi-elasticities/elasticities: ˆp, ˆw defined as percent responses per unit μ; wage-gap vector wgap = ˆw_L − ˆw_H.
- Functional-form assumptions:
  - Baseline Cobb-Douglas technology and preferences (elasticity of substitution = 1) to obtain Φ_p and Φ_μ from data. Authors note direction of impacts robust to functional-form choice; magnitudes vary with substitution elasticities.
- Relative wage decomposition:
  - ˆp^n_j = Σ_m Σ_i θ^{mn}_{ij} ˆp^m_i + (θ_L)^n_j ˆw^n_L + (θ_H)^n_j ˆw^n_H + (θ_K)^n_j ˆr^n_j.
  - Wage-gap decomposition (equation (23)/(40)) isolates effects via supply prices and via input prices, split by domestic tradables, domestic nontradables, and imports.

### Data and calibration
- Data source: 1995 Asian international input-output table (IDE (2001)), aggregated to 10 industries per country.
- Sample: Indonesia, Malaysia, Philippines, Singapore, Thailand, China, Taiwan Province of China, Korea (Republic of), Japan, and the United States; Hong Kong SAR treated as part of the rest of the world.
- Industry aggregation: 78 industries → 10 industries (labels include AGR, EGW, FOD, TEX, PAP, CHE, BMI, MEQ, SER, MOT).
- Tariff and transport data:
  - Transport cost rates d^mn estimated using U.S. bilateral trade freight estimates (Hummels (1999), Feenstra et al. (2002)) via regressions on distance.
  - MFN tariff rates t^n_i from TRAINS (UNCTAD); averaged across commodities to get ˆt^n_i and applied to bilateral flows to build T^mn_ij; adjustments to match aggregate actual tariff payments where needed.
- Wage and employment data:
  - Wage data from Freeman and Oostendorp (2000); employment by occupation and industry from national labor statistics; constructed relative wage compensation of production workers.
- Baseline numeric facts (1995 within-region trade and tariffs):
  - Total amount traded in the region: $680.071 billion.
  - Amount of import duties incurred (within-region): $40.854 billion.
  - Corresponding tariff rate (region average): 6.0 percent.
- Numeraire: Total exports to the rest of the world fixed in nominal terms (p_m f_{m,M+1} fixed); real exports F_{m,M+1} endogenous.

### Empirical findings — direct effects (initial excess supply/demand shifts)
- Abolishing tariffs increases supply, intermediate demand, and consumption demand in every goods market.
- Aggregate excess supply across the region: $10.841 billion (sum across 100 goods markets) → downward pressure on general price levels.
- Country-level aggregate excess demand (sign convention: negative = excess demand):
  - Singapore: −$0.740 billion.
  - China: −$2.225 billion.
  - United States: −$1.645 billion.
- Sectoral pattern:
  - Excess demand concentrated in nontradables:
    - Agriculture aggregate excess demand: $3.777 billion.
    - Services aggregate excess demand: $23.242 billion.
  - Excess supply concentrated in tradables:
    - MEQ (machinery/equipment) aggregate excess supply: $21.098 billion.
    - MOT (other industrial products) aggregate excess supply: $12.199 billion.
- Interpretation: Abolishing existing tariffs induces expansion in tradables and contraction in nontradables, reallocating resources across sectors.

### Empirical findings — effect on prices
- Price adjustments computed via Γ_μ = −Φ_p^{−1} Φ_μ; Paasche price index used as country general price level measure.
- Aggregate/representative results (Table 5, "Total" row, upper block):
  - Average change in total price level (region average): −0.725 percent.
- Country-level Paasche price index changes (Total row, upper block, in percent):
  - IDN −1.298
  - MYS −2.121
  - PHL −3.361
  - SGP −0.768
  - THA −4.341
  - CHN 0.649
  - TWN −0.977
  - KOR −1.684
  - JPN −1.227
  - USA −0.300
- Relative prices (good j relative to country price index):
  - Relative prices of nontradables rise in all countries.
  - Relative prices of tradables generally fall in all countries.
- Interpretation: Price declines associated with excess supply are larger for countries with larger tariff cuts; China’s general price level increases while most others fall (largest declines in THA −4.341 percent and PHL −3.361 percent).

### Empirical findings — effect on trade (direct + indirect)
- Direct effects (before price adjustment):
  - Regional trade volume initially increases by $20.386 billion.
  - Domestic demand initially increases by $80.262 billion.
- Indirect effects (after price adjustment):
  - Price changes reduce the initial trade increase: after adjustment regional trade volume increase = $14.628 billion.
  - Demand for domestic goods after adjustment increases by $32.071 billion.
- Country-level examples (Table 7; component breakdowns reported in source):
  - Thailand: reported real total effects include components 2.604 (domestic), 1.965 (export), −0.355 (import) in table entries.
  - China: components 4.149 (domestic), 2.021 (export), 1.180 (import) reported.
  - United States: components 4.875 (domestic), 1.488 (export), 5.385 (import) reported.
- Patterns:
  - Countries with relative price falls (e.g., THA, PHL) see increases in export demand and improvements in real trade balance.
  - Countries with relative price increases (e.g., CHN, USA) see decreases in export demand and deterioration in real trade balance.
  - Overall substitution toward foreign goods raises trade volume proportionally more than domestic demand: trade volume increase 1.79 percent vs domestic goods demand increase 0.26 percent (relative to initial levels).

### Empirical findings — employment and factor reallocation
- Labor demand percent changes (Table 8, "Total" row):
  - Less-skilled labor percent changes:
    - IDN 0.144
    - MYS 0.304
    - PHL 0.246
    - SGP 0.424
    - THA 0.233
    - CHN −0.048
    - TWN 0.207
    - KOR 0.232
    - JPN 0.035
    - USA 0.053
    - Total 0.185
  - Skilled labor percent changes:
    - IDN 0.112
    - MYS 0.301
    - PHL 0.223
    - SGP 0.423
    - THA 0.315
    - CHN −0.063
    - TWN 0.202
    - KOR 0.248
    - JPN 0.045
    - USA 0.061
    - Total 0.189
  - Relative change in demand for less-skilled labor (Total row, summary):
    - IDN 0.032
    - MYS 0.003
    - PHL 0.023
    - SGP 0.001
    - THA −0.082
    - CHN 0.015
    - TWN 0.006
    - KOR −0.016
    - JPN −0.011
    - USA −0.008
    - Total −0.004
- Patterns:
  - Emerging markets often see larger increases in less-skilled labor demand in nontradables and declines in tradables.
  - Industrialized economies (JPN, USA) tend to see larger increases in skilled labor demand than less-skilled, implying relative decreases in demand for less-skilled labor.

### Empirical findings — effect on wages and decomposition
- Change in real wage rates and wage-gap (Table 9; all entries in percent):
  - Real wage changes and wage gap (less-skilled, skilled, wage gap = less-skilled − skilled):
    - IDN: 0.442; 0.475; −0.032.
    - MYS: 1.473; 1.476; −0.003.
    - PHL: 1.576; 1.600; −0.023.
    - SGP: 0.551; 0.552; −0.001.
    - THA: 1.739; 1.657; 0.082.
    - CHN: 0.355; 0.370; −0.015.
    - TWN: 0.697; 0.702; −0.006.
    - KOR: 0.619; 0.603; 0.016.
    - JPN: 0.165; 0.154; 0.011.
    - USA: 0.056; 0.048; 0.008.
- Decomposition of wage-gap changes:
  - Supply-price channel: tradables and nontradables — generally negative (supply-price declines tend to widen wage gap against less-skilled).
  - Input-price channel: domestic tradables, domestic nontradables, imports — domestic input price declines generally raise demand for less-skilled labor (narrow wage gap); imported input-price effects depend on terms of trade.
  - Countries with deteriorating terms of trade (large tariff cuts and large domestic price falls) face higher relative import input costs → negative effect on wage gap.
  - Domestic tradable input-price declines often offset adverse terms-of-trade effects; net effect on wage gap is typically small.
- Key aggregate finding:
  - Abolishing the existing 1995 within-region tariff rates (average 6.0 percent) yields a negligible effect on the wage gap across the region because offsetting supply-price and input-price channels operate simultaneously.

### Sensitivity analysis and robustness
- Functional-form sensitivity:
  - Baseline Cobb-Douglas (es = 1); alternative CES elasticities examined (es = 1.5, es = 3).
  - Direction of trade-balance impacts robust to changes in elasticity of substitution; magnitudes increase with higher elasticity.
  - Figure 2 (in source) illustrates sensitivity: qualitative patterns preserved; magnitudes larger with higher es.

### Policy-relevant interpretations and implications
- Trade policy impacts on wages are mediated importantly by price-feedback effects via world/regional prices when member countries are large enough to affect those prices.
- Net effect of trade liberalization on less-skilled vs skilled wage gap can be small even when terms-of-trade effects are nontrivial because:
  - Supply-price declines (negative for less-skilled relative wages) can be offset by input-price declines (positive for less-skilled relative wages), especially via domestic tradable input cost reductions.
- Evaluation of trade liberalization impacts should:
  - Model region-wide general equilibrium price feedbacks (not assume each country is small).
  - Explicitly include intermediate input trade and its role in transmitting tariff changes to input costs and wages.
  - Disaggregate tradables vs nontradables and intermediate vs final trade to capture substitution and demand/supply responses.
- Empirical policy takeaway: Abolishing the 1995 within-region tariffs (average 6.0 percent) yields non-negligible macro and sectoral effects (price level, trade reallocation, employment shifts), but a negligible average effect on the wage gap across the region because offsetting channels operate simultaneously.

*Source: _wp06155 — 2. SensitivityAnalysis:TradeBalance (excerpted content from the provided IMF PDF chapter/section).*

### References............................................... 37

### _wp06155 - References............................................... 37

### Tables
- 1.   Countries and Industries  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  13
- 2.   Basic Accounting Scheme: International Input-Output Table . . . . . . . . . . . .  14
- 3.   Import Duties and TariffRatesonWithin-RegionTrade,byCountry .......... 16
- 4.   Direct Effect on Excess Supply . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  17
- 5.  ChangeofPrice ........................................ 19
- 6.  ChangeofDomesticDemandandRegionalTrade,byCountry .............. 20
- 7.  ChangeofTradeBalance ................................... 21
- 8.   Change of Labor Input  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  23
- 9.  SourcesofChangeinWageGap ............................... 24
- 10. SummaryofInverseofJacobianMatrix........................... 29
- 11. RelativeWageRatesofNonproductionWorkers,byIndustry............... 35
- 12.  Relative Wage Compensation of Production Workers, by Industry  . . . . . . . . .  35
- 13. SourceofEmploymentData ................................. 36

### Figures
- 1.   Jacobian MatrixΦ
p
...................................... 11

*Source: _wp06155 - References............................................... 37*

### 2.  SensitivityAnalysis:TradeBalance ............................. 31

### 2.  SensitivityAnalysis:TradeBalance

### Introduction and research question
- Literature review: prior work (Harrigan (2000), Harrigan and Balaban (1999), Tokarick (2005), Feenstra and Hanson (2001)) finds domestic factors (relative domestic prices, relative factor supply, nontradables) are primary drivers of increased wage inequality; foreign price changes found, at most, small direct effects on relative wages.
- Central question: Do trade policies (tariff changes, FTAs) have negligible effects on domestic relative wages if they affect world prices? Effect on domestic relative wages depends on whether trade policy affects domestic prices via world prices.
- Paper advances:
  - Constructs a general equilibrium model for a “region” where member countries are large enough to affect regional world prices.
  - Uses 1995 international input-output data for 9 APEC member countries (plus associated regions), distinguishing tradables vs nontradables and final vs intermediate trade.
- Key simulation: abolishing existing tariff payments (setting observed 1995 within-region tariff rates to zero).

### Model and methodology
- Structure:
  - M number of countries, I number of industries; (M+1)th country = rest of the world.
  - Representative producer in industry j in country n: production function X^n_j = f^n_j(x^1n_j,...,x^Mn_j, L^n_j, H^n_j; x̄^{M+1,n}_j, K̄^n_j) with x̄ and K̄ fixed.
  - Profit maximization with import prices p^m_n (local prices include tariff and transport costs): p^m_i (1 + τ^mn_i) with (1 + τ^mn_i) = (1 + d^mn_i)(1 + t^mn_i).
  - Household: utility U_n(c^1n,...,c^Mn), consumption expenditure C_n = δ_n(p_n x_n − Σ_j Σ_m p^m_n' x^{m n}_j).
  - Competitive equilibrium: prices (p^1,...,p^M, w_L, w_H) and allocation satisfying firms’ profit maximization, households’ utility maximization, and market clearing for goods and two labor types.
- Linearization and comparative statics:
  - Define excess supply functions e and Jacobian Φ_p of excess supply w.r.t. prices and wages.
  - Small change in shift parameters μ (percentage change in tariff rates) satisfies Φ_p · Γ_μ + Φ_μ = 0, hence Γ_μ = −Φ_p^{−1} Φ_μ.
  - Semi-elasticities/elasticities: ˆp, ˆw are ∂p/p ∂μ etc.; wage-gap vector wgap = ˆw_L − ˆw_H.
- Functional-form assumptions for estimation:
  - Cobb-Douglas technology and preferences (elasticity of substitution = 1) used to obtain Φ_p and Φ_μ from data. Authors note direction of impacts robust to functional-form choice; magnitudes vary with substitution elasticities.
- Relative wage decomposition:
  - Price change for good j in country n decomposed: ˆp^n_j = Σ_m Σ_i θ^{mn}_{ij} ˆp^m_i + (θ_L)^n_j ˆw^n_L + (θ_H)^n_j ˆw^n_H + (θ_K)^n_j ˆr^n_j with weights summing to 1.
  - Wage-gap decomposition formula (equation (23) / (40)) isolates effects via supply prices (prices of goods sold) and via input prices (intermediate inputs), and further splits by domestic tradables, domestic nontradables, and imports.

### Data and calibration
- Data source: 1995 Asian international input-output table (IDE (2001)), aggregated to 10 industries per country; sample includes Indonesia, Malaysia, Philippines, Singapore, Thailand, China, Taiwan Province of China, Korea (Republic of), Japan, and the United States; Hong Kong SAR treated as part of rest of the world.
- Industry aggregation: 78 industries → 10 industries due to labor data constraints. Industry labels (examples): AGR, EGW, FOD, TEX, PAP, CHE, BMI, MEQ, SER, MOT.
- Tariff and transport data:
  - Transport cost rates d^mn estimated using U.S. bilateral trade freight estimates (Hummels (1999), Feenstra et al. (2002)) via regressions on distance.
  - MFN tariff rates t^n_i obtained from TRAINS (UNCTAD) and concorded to UIO classifications; averaged across commodities to get ˆt^n_i and applied to bilateral flows to build T^mn_ij (adjustment applied to match aggregate actual tariff payments where needed).
- Skilled vs less-skilled wages and employment:
  - Wage data from Freeman and Oostendorp (2000) (October Inquiry, ILO); employment by occupation and industry from national labor statistics; constructed relative wage compensation of production workers (relative to nonproduction workers).
- Key baseline numeric facts (1995 within-region trade and tariffs):
  - Total amount traded in the region: $680.071 billion.
  - Amount of import duties incurred (within-region): $40.854 billion.
  - Corresponding tariff rate (region average): 6.0 percent.
- Numeraires and exports to rest of world:
  - Total exports to the rest of the world treated as numeraire and fixed in nominal terms (p_m f_{m,M+1} fixed), so real exports F_{m,M+1} are endogenous.

### Empirical findings — direct effects (immediate shifts in supply/demand)
- Abolishing tariffs increases supply, intermediate demand, and consumption demand in every goods market (Table 4 summary).
- Aggregate excess supply across the region: $10.841 billion (sum across 100 goods markets), implying downward pressure on general price levels.
- Country-level aggregate excess demand (instead of excess supply):
  - Singapore: excess demand −$0.740 billion (i.e., excess demand of $0.740 billion shown as negative sign in table).
  - China: excess demand −$2.225 billion.
  - United States: excess demand −$1.645 billion.
- Sectoral pattern:
  - Excess demand concentrated in nontradables (agriculture and services): aggregate excess demand sums $3.777 billion (agriculture) and $23.242 billion (services).
  - Excess supply concentrated in tradables (machinery and other industrial products): aggregate excess supply sums $21.098 billion (MEQ) and $12.199 billion (MOT).
- Interpretation: abolishing existing tariffs induces expansion in tradables, contraction in nontradables, and resource reallocation across sectors.

### Empirical findings — effect on prices
- Price adjustments Γ_μ computed via Γ_μ = −Φ_p^{−1} Φ_μ; Paasche price index used to represent general price level change per country.
- Aggregate/Representative results (Table 5):
  - Average change in total price level (last row, upper block, “Total”): −0.725 percent (region average).
  - Country-level Paasche price index changes (Total row, upper block): IDN −1.298, MYS −2.121, PHL −3.361, SGP −0.768, THA −4.341, CHN 0.649, TWN −0.977, KOR −1.684, JPN −1.227, USA −0.300 (all in percent).
  - Notable: China’s general price level increases (0.649 percent); most other countries see declines, largest in Thailand (−4.341 percent) and the Philippines (−3.361 percent).
- Relative prices (good j relative to country price index; lower block of Table 5):
  - Relative prices of nontradables rise in all countries; relative prices of tradables generally fall in all countries.
- Interpretation: price declines associated with excess supply are larger for countries with larger tariff cuts; general price level falls where tariff reductions are large and demand is predominantly intraregional.

### Empirical findings — effect on trade (direct + indirect)
- Direct effects (initial increases prior to price adjustment): regional trade volume initially increases by $20.386 billion; domestic demand increases by $80.262 billion (Table 6 first block).
- Indirect effects (price adjustments):
  - Price changes reduce the initial trade increase: total (after adjustment) regional trade volume increase = $14.628 billion (Table 7).
  - Demand for domestic goods after adjustment increases by $32.071 billion (Table 7).
- Country-level real trade balance total effects (Table 7, selected entries):
  - Thailand real total effect on trade balance: 2.604 (domestic) 1.965 (export) −0.355 (import) → nominal/real trade balance changes reported (see table for full breakdown).
  - China: real total effect shows positive real trade balance change equivalent components 4.149 (domestic) 2.021 (export) 1.180 (import) → net effects described in table.
  - United States: domestic 4.875, export 1.488, import 5.385 → trade balance effects reported in table.
- Patterns:
  - Countries with relative price falls (e.g., Thailand, Philippines) see increases in export demand and improvements in real trade balance.
  - Countries with relative price increases (e.g., China, United States) see decreases in export demand and deterioration in real trade balance.
  - Overall: initial increase in trade volume (direct effect) is partly offset by price adjustments (indirect effect); substitution toward foreign goods raises trade volume proportionally more than domestic demand (trade volume increase 1.79 percent vs domestic goods demand increase 0.26 percent, relative to initial levels).

### Empirical findings — employment and factor reallocation
- Labor demand changes (Table 8):
  - Total effect on less-skilled labor (Total row, upper block) percent changes: IDN 0.144, MYS 0.304, PHL 0.246, SGP 0.424, THA 0.233, CHN −0.048, TWN 0.207, KOR 0.232, JPN 0.035, USA 0.053, Total 0.185 (percent).
  - Total effect on skilled labor (Total row, lower block) percent changes: IDN 0.112, MYS 0.301, PHL 0.223, SGP 0.423, THA 0.315, CHN −0.063, TWN 0.202, KOR 0.248, JPN 0.045, USA 0.061, Total 0.189 (percent).
  - Relative increase in demand for less-skilled labor (Total row, lower summary): IDN 0.032, MYS 0.003, PHL 0.023, SGP 0.001, THA −0.082, CHN 0.015, TWN 0.006, KOR −0.016, JPN −0.011, USA −0.008, Total −0.004 (percent).
- Patterns:
  - Emerging market economies commonly see larger increases in less-skilled labor demand in nontradables and declines in tradables.
  - Industrialized economies (Japan, USA) tend to see larger increases in skilled labor demand than less-skilled, implying relative decreases in demand for less-skilled labor.

### Empirical findings — effect on wages and decomposition
- Change in real wage rates and wage-gap (Table 9):
  - Change in real wage rate less-skilled (column 1) and skilled (column 2) and wage gap (column 3 = less-skilled − skilled) for selected countries (all in percent):
    - IDN: less-skilled 0.442; skilled 0.475; wage gap −0.032.
    - MYS: 1.473; 1.476; −0.003.
    - PHL: 1.576; 1.600; −0.023.
    - SGP: 0.551; 0.552; −0.001.
    - THA: 1.739; 1.657; 0.082.
    - CHN: 0.355; 0.370; −0.015.
    - TWN: 0.697; 0.702; −0.006.
    - KOR: 0.619; 0.603; 0.016.
    - JPN: 0.165; 0.154; 0.011.
    - USA: 0.056; 0.048; 0.008.
  - Sources of change in wage gap split into:
    - Supply prices: tradables and nontradables (columns 4 and 5) — generally negative (wage gap widens when supply prices fall).
    - Input prices: domestic tradables, domestic nontradables, and imports (columns 6–8) — domestic input price declines generally increase demand for less-skilled labor (positive effect on wage gap); imported input price effects depend on terms of trade.
- Key aggregate findings:
  - Abolishing the existing 1995 within-region tariff rates (average 6.0 percent) has a negligible effect on the wage gap across countries (wage-gap column values are small, often near zero).
  - Supply-price channel: tariff cuts create excess supply → lower supply prices → negative effect on less-skilled relative wages (wage-gap widens).
  - Input-price channel: tariff cuts lower input prices (especially tradable inputs) → lower production costs → expand output and demand for less-skilled labor → positive effect on wage-gap (wage-gap narrows).
  - Import input-price channel: countries with deteriorating terms of trade (large tariff cuts and large domestic price falls) face higher relative import input costs → negative effect on wage-gap; countries with improving terms of trade benefit (positive effect).
  - In many cases, the domestic input-price effects (especially domestic tradable inputs) are large enough to offset the adverse terms-of-trade effects; net effect on wage gap is small/near zero.

### Sensitivity analysis and robustness
- Functional-form sensitivity:
  - Cobb-Douglas (es = 1) baseline; alternative CES elasticities examined.
  - Direction of trade-balance impacts robust to changes in elasticity of substitution; magnitudes increase with higher elasticity.
  - Figure 2 illustrates sensitivity of trade balance by country for es = 1, es = 1.5, es = 3 (qualitative pattern preserved; magnitudes larger with higher es).

### Policy-relevant interpretations and implications
- Trade policy impact on wages is mediated importantly by price-feedback effects via world/regional prices when member countries are large enough to affect those prices.
- The net effect of trade liberalization on the less-skilled vs skilled wage gap can be small even when terms-of-trade effects are nontrivial, because:
  - Supply-price declines (negative for less-skilled relative wages) can be offset by input-price declines (positive for less-skilled relative wages), especially via domestic tradable input cost reductions.
- Evaluation of trade liberalization impacts should:
  - Model region-wide general equilibrium price feedbacks (not assume each country is small).
  - Explicitly include intermediate input trade and its role in transmitting tariff changes to input costs and wages.
  - Disaggregate tradables vs nontradables and intermediate vs final trade to capture substitution and demand/ supply responses.
- Empirical policy takeaway: Abolishing the 1995 within-region tariffs (average 6.0 percent) yields non-negligible macro and sectoral effects (price level, trade reallocation, employment shifts), but a negligible average effect on the wage gap across the region because offsetting channels operate simultaneously.

*Source: _wp06155 - 2.  SensitivityAnalysis:TradeBalance (excerpted content from the provided IMF PDF chapter/section).*

### References

### References

### Listed references
- Armington, Paul S., 1969, “A Theory of Demand for Products Distinguished by Place of Produc-
tion,”StaffPapers, International Monetary Fund,Vol. 16, No. 1, pp. 170—201.
- Feenstra, Robert C., and Gordon Hanson, 2001, “Global Production Sharing and Rising Inequality:
A Survey of Trade and Wages,” inHandbook of International Trade, ed. by Kwan Choi and
James Harrigan (Malden, Massachusetts: Blackwell).
- Feenstra, Robert C., John Romalis, and Peter K. Schott, 2002, “U.S. Imports, Exports, and Tariff
Data, 1989-2001,” NBER Working Paper No. 9387 (Cambridge, Massachusetts: National
Bureau of Economic Research).
- Feltenstein, Andrew, and Florenz Plassmann,2005, “The Welfare Analysis of a Free Trade Zone:
Intermediate Goods and the Asian Tigers,” MAS StaffPaper No. 41 (Singapore: Monetary
Authority of Singapore).
- Freeman, Richard B., and Remco H. Oostendorp, 2000, “Occupational Wages Around the World:
Pay Across Occupations and Countries,” NBER Working Paper No. 8058 (Cambridge, Massa-
chusetts: National Bureau of Economic Research).
- Harrigan, James, 2000, “International Trade and American Wages in General Equilibrium, 1967—
1995,” inThe Impact of International Trade on Wages, ed. by Robert C. Feenstra (University
of Chicago Press).
- –––, and Rita A. Balaban, 1999, “U.S. Wage Effects in General Equilibrium: The Effects
of Prices, Technology, and Factor Supplies, 1963—1991,” NBER Working Paper No. 6981
(Cambridge, Massachusetts: National Bureau of Economic Research).
- Hummels, David, 1999, “Toward a Geography of Trade Costs” (unpublished, West Lafayette,
Indiana: Purdue University).
- Institute of Development Economies (IDE), 2001,Asian International Input-Output Table 1995,
IDE Statistical Data Series No. 82, (Tokyo: IDE-JETRO).
- Katz, Lawrence F., and David Autor, 1999, “Changes in the Wage Structure and Earnings In-
equality,” inHandbook of Labor Economics,Vol. 3A, ed. by Orley Ashenfelter and David
Card (Amsterdam: Elsevier).
- Lawrence, Robert Z., 1994, “Trade, Multinationals, and Labor,” NBER Working Paper No. 4836
(Cambridge, Massachusetts: National Bureau of Economic Research).
- –––, and Matthew J. Slaughter, 1993, “International Trade and American Wages in the 1980s:
Giant Sucking Sound or Small Hiccup?”Brookings Papers on Economic Activity: 2, Brookings
Institution, pp. 161—226.
- McKenzie, Lionel W., 1960, “The Matrix with Dominant Diagonal and Economic Theory,” in
Mathematical Methods in the Social Sciences, ed. by Kenneth J. Arrow, Samuel Karlin, and
Patrick Suppes (Palo Alto, California: Stanford University Press).
- -38-
- Tokarick, Stephen P., 2005, “Quantifying the Impact of Trade on Wages: The Role of Nontraded
Goods,”Review of International Economics, Vol. 13, pp. 841—60.
- Tokutsu, Ichiro, 2002, “Quadrilateral Market Dependence: Japan, USA, European Union, and
Asian Countries,”Journal of Policy Modeling, Vol. 23, pp. 119—50.
- United Nations Conference on Trade and Development (UNCTAD), 1996,Trade Analysis and
Information System (TRAINS),Version 3.0 (Geneva).

*Source: _wp06155 - References*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06155.pdf_