## _wp06147

## Source details

**Canonical URL:** [_wp06147](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06147.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06147.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06147.pdf.json)

---

### 1. Data Description and Sources
- Focus: empirical relationship between government expenditures and imports to assess effects of fiscal policy on the trade account.
- Sample: annual panel data of the G-7 countries for the years 1970 through 2002 (services series from 1977 through 2002 where needed).
- Estimation method: pooled mean group estimation (PMG) allowing intercepts, short-run coefficients, and error variances to differ across countries while constraining long-run coefficients to be the same.
- Demand components disaggregated: private consumption (C), private sector investment (I), government expenditure (G), and exports (X).
- Key data sources and variable definitions (preserved terminology):
  - XG — Goods export volumes: "Export volumes (IFS line 72) ... using the 1995 average for merchandise exports in US$ (IFS line 78aa) ... deflating by PC."
  - XS — Service export volumes: "Service credits in US$ (IFS line 78ad) ... deflating by PC."
  - MG — Domestic goods import volumes: "import volume FOB series (IFS line 73) ... merchandise exports in US$ (IFS line 78ab) ... using the 1995 average for the US$ exchange rate (r)."
  - MS — Domestic service import volumes: "Service debits in US$ (IFS line 78ae) ... deflating by PCW after converting PCW into domestic currency terms using EFEX."
  - YG* — World income for goods exports: "total world exports in US$ at current prices (IFS line 70), deflated using WPXG."
  - YS* — World income for service exports: "Total OECD GDP at constant market prices in US$."
  - Y — Domestic real GDP: "IFS line 99b and deflated by PY."
  - C — real private consumption: "IFS line 96f and deflated by CP."
  - I — Real private sector investment: "IFS line 93i plus IFS line 93e and deflated by PY."
  - G — real government expenditure: "IFS line 91f and deflated by PY."
  - X — real exports: "IFS line 90c and deflated by PY."
  - Price and exchange rate measures: PC (IFS line 64), PCW (MEI of the OECD), PXG (IFS line 76), WPXG (IFS line 74), PY (IFS line 99bi), r (IFS line rf), EFEX (constructed from r and DOTS).
  - Relative prices: RPXG = (WPXG*r)/PXG; RMPG = (WPXG*r)/PD; RPS = PCW/(PC*EFEX).

### II. The Government Sector and the Trade Account (cross-country evidence)
- Cross-country evidence (selected import contents, Table 1 excerpt):
  - Aggregate expenditure: 0.243, 0.198, 0.216, 0.235, 0.200
  - Private consumption: 0.264, 0.208, 0.229, 0.249, 0.200
  - Government expenditure: 0.134, 0.060, 0.064, 0.097, 0.132
  - Gross investment: 0.244, 0.267, 0.261, 0.372, 0.318
  - Exports: 0.272, 0.201, 0.241, 0.235, 0.224
- Theoretical proposition: "the smaller the size of the government—measured as government expenditures in percent of GDP—the higher the import-to-GDP ratio, because government expenditure has a smaller import content than private consumption."
- Empirical scatterplots (G-7 annual data 1990–2004) and fitted lines (examples):
  - General government panels: y = -0.90x + 0.44; y = -0.56x + 0.53; y = -1.04x + 0.84; y = 0.17x + 0.03 (t-statistics shown beneath coefficients).
  - Central government panels: y = -1.34x + 0.33; y = 0.42x + 0.02; y = -1.83x + 0.72; y = -0.86x + 0.43 (t-statistics shown).
  - With the exception of Japan, coefficients are significant at the 1 percent level. Japan shows a positive association (attributed to its prolonged stagnation).
- Cross-country correlations (1990–2004, selected, significance reported):
  - Canada: -0.88***
  - Germany: -0.57***
  - United Kingdom: -0.59***
  - Italy: -0.94***
  - Japan: 0.42*
  - United States: -0.84***
  - Austria: -0.81***
  - Belgium: -0.89***
  - Sweden: -0.96***
  - Switzerland: 0.06
  - Norway: 0.43*
  - New EU Member States examples: Cyprus -0.98***; Hungary -0.73***; Malta -0.68***; Slovak Republic -0.62***.
- Note: correlations indicate a robust negative relationship in many countries but do not quantify causal impact of government expenditure changes on imports or the current account.

### III. Model Specification
- Standard trade model (log-form) relates exports and imports to domestic income (Y), foreign income (Y*), and relative prices (RP).
- Rationale for disaggregation: allow distinct import elasticities for ln C, ln I, ln G, and ln X (extended goods imports equation (8); extended service imports equation (9)).
- Theoretical expectations:
  - Income elasticities close to 1 but may deviate empirically.
  - Sum of demand elasticities (C, I, G, X) should be equal or close to 1 in extended equations.
  - Relative price coefficients expected negative.

### IV. Empirical Analysis and Results
- Unit root and cointegration:
  - Panel unit root tests applied: Levin, Lin, and Chu (LLC) and Breitung. Almost all variables integrated of order one.
  - Panel cointegration tests: Pedroni and Kao reject the null of no cointegration in all trade volume equations.
  - Example unit root test p-values (Table 3 highlights):
    - RPXG: LLC 0.0579; Breitung 0.0584
    - RMPG: LLC 0.8771; Breitung 0.4112
    - YS*: LLC 0.0001; Breitung 0.9760
    - C: LLC 0.0000; Breitung 0.8933
    - G: LLC 0.0001; Breitung 0.2520
- Conventional trade volume PMG estimates (Table 5, long-run coefficients, significance ** 1 percent):
  - Goods export price elasticity: -0.849** (t = -8.647)
  - Goods export income elasticity: 0.906** (t = 36.395)
  - Service export price elasticity: -0.726** (t = -3.500)
  - Service export income elasticity: 1.018** (t = 3.572)
  - Goods import price elasticity: -0.313** (t = -3.076)
  - Goods import income elasticity: 1.953** (t = 9.896)
  - Service import price elasticity: -1.263** (t = -15.921)
  - Service import income elasticity: 1.316** (t = 56.190)
  - Joint Hausman tests (examples): 0.66, 0.89, 0.31, 0.94 (null of slope homogeneity not rejected).
- Extended import volume PMG estimates (Table 6, long-run coefficients; significance: * 5 percent, ** 1 percent):
  - Goods imports (extended):
    - Price elasticity: -0.665** (-5.015)
    - Private consumption (ln C): 1.102** (3.481)
    - Government expenditure (ln G): 0.392* (1.762)
    - Private sector investments (ln I): 0.427** (5.152)
    - Exports (ln X): 0.435** (4.156)
    - Joint Hausman test: 0.12
  - Service imports (extended):
    - Price elasticity: -1.592** (-6.747)
    - Private consumption (ln C): 1.433** (1.916)
    - Government expenditure (ln G): 0.491** (2.485)
    - Private sector investments (ln I): 0.030 (0.076) [not significant]
    - Exports (ln X): 0.503** (1.972)
    - Joint Hausman test: 0.22
- Key empirical findings:
  - Demand composition matters: elasticities differ across private consumption, private investment, government expenditure, and exports.
  - Government expenditure has the smallest import elasticity among demand components but is positive and significant.
  - A lasting increase in government expenditure of 1 percent leads to:
    - an increase of goods imports of 0.4 percent (0.392* reported).
    - an increase of service imports of 0.5 percent (0.491** reported).
  - Ceteris paribus implication: an increase in government expenditure leads to a deterioration of the trade account.
- Caveats on indirect effects:
  - Ceteris paribus interpretation may be misleading if fiscal changes induce crowding out or crowding in of private demand.
    - Prior findings: Blanchard and Perotti (2002) — fiscal expansion raises consumption and lowers investment.
    - Fatás and Mihov (2001) — consumption increases after expenditure shocks; investment not significantly affected.
    - Karras (1994) — private consumption and government spending complementary; consumption decreases when government expenditures are cut.
  - If government spending crowds out private investment but raises private consumption, the net effect on imports depends on magnitudes and elasticities.
  - For services: private investment has negligible effect; government and private consumption increases raise service imports.
  - Because the goods account is larger than the service account, goods-account effects likely dominate overall trade-account outcomes.

### V. Summary and Conclusion
- Main conclusions:
  - Fiscal policy matters for the trade account through direct and indirect channels.
  - Disaggregating demand reveals that government and private demand components have different import elasticities; assuming common elasticities can be misleading.
  - PMG panel estimation for G-7 (1970–2002) finds that an increase in government expenditures by 1 percent leads to:
    - an increase in goods imports of about 0.4 percent.
    - an increase in service imports of almost 0.5 percent.
  - Policy implication (ceteris paribus): an increase in government expenditure would lead to a deterioration of the trade account.
- Policy-relevant interpretation and recommendation:
  - The overall impact of fiscal expansions or contractions on the current account depends on:
    - (a) the relative import elasticities of demand components, and
    - (b) how private consumption and investment respond to fiscal changes.
  - Given empirical uncertainty about private-sector responses to fiscal policy, policymakers should be cautious in assuming a simple one-to-one relationship between changes in government spending and trade-account outcomes.
- Suggested further research:
  - "Further research could determine the overall impact (i.e. the direct impact of a change in expenditure and the indirect impact through the reaction of private demand) that a change in government expenditure could have on the trade account of a particular country."
  - Recommended approach: country-specific analysis of the link between fiscal policy measures and private demand.

### Conclusions if an increase (decrease) in government expenditure was to crowd out (crowd in)
- Ambiguity:
  - An increase (decrease) in government expenditure could crowd out (crowd in) private demand; "If this crowding-in/out effect is strong enough, an increase in government expenditures could bring about the opposite result."
  - The net effect depends on whether crowding out predominates (potential improvement in the trade account if private demand with higher import content is displaced) or crowding in predominates (worsening of the trade account).
- Evidence and magnitudes:
  - Service imports are less than one-third of the size of goods imports in the G-7 context, implying goods-account responses likely dominate overall outcomes.
  - Literature documents divergent private responses to fiscal shocks (examples cited in the study).
- Policy implication:
  - To determine the overall impact of fiscal changes on the trade account, one must combine the direct import response to government expenditure with estimated private-sector reactions (consumption and investment) for the country of interest.

*Source: _wp06147*

### 1.  Data Description and Sources

### 1.  Data Description and Sources

### Introduction and research question
- Focus: empirical relationship between government expenditures and imports to assess effects of fiscal policy on the trade account.
- Motivation: existing reduced-form studies produce ambiguous results; structural modeling and long-run time-series techniques are required to reveal underlying causalities.
- Key insight: imports are driven by domestic demand factors, while exports depend on external demand factors; isolating government expenditure effects requires disaggregation of demand.

### Model specification and identification strategy
- Demand components included separately: private consumption, private sector investment, government expenditure, and exports.
- Rationale: different components exhibit different import elasticities; government consumption generally has lower import content than other demand components.
- Contrast with conventional trade equations: conventional specifications use total demand as the explanatory variable; this study estimates import equations on disaggregate demand variables (private consumption, private sector investment, government expenditure, exports).

### Data and estimation technique
- Sample: annual panel data of the G-7 countries for the years 1970 through 2002.
- Estimation method: pooled mean group estimation.
  - Allows intercepts, short-run coefficients, and error variances to differ freely across countries.
  - Constrains the long-run coefficients to be the same for all cross-sections.
- Purpose of technique: account for cross-country differences while identifying common long-run relationships between import volumes and demand components.

### Main empirical findings and interpretation
- Finding: a change in government expenditure has a significant positive impact on both goods and service imports.
  - Implication: ceteris paribus, an increase in government expenditure would lead to a deterioration of the trade account.
- Important caveat: the ceteris paribus interpretation may be misleading if government expenditure changes induce crowding out or crowding in of private demand components.
  - If government spending crowds out private demand sufficiently, an increase in government expenditures could lead to an improvement in the trade account (the opposite result).
  - Therefore, net effect depends on strength and direction of crowding in/out effects across private consumption and investment.

### Structural and policy relevance
- Structural modeling of the trade account is crucial to trace causal channels from fiscal policy to external balances.
- Disaggregating demand components into private versus public is essential because public and private demand have different import contents and elasticities.
- Policy conclusions require consideration of general equilibrium responses (private demand adjustments) rather than ceteris paribus partial-equilibrium effects.

*Source: _wp06147 - 1.  Data Description and Sources*

### conclusions.

### conclusions.

### II. THE GOVERNMENT SECTOR AND THE TRADE ACCOUNT
- Cross-country evidence (Germany, France, the United Kingdom, Italy, and the United Kingdom separately for 2001) shows differing import contents across demand components; government expenditure reveals the lowest import content across countries (Table 1).
- Table 1 (excerpted import contents):
  - Aggregate expenditure: 0.243, 0.198, 0.216, 0.235, 0.200
  - Private consumption: 0.264, 0.208, 0.229, 0.249, 0.200
  - Government expenditure: 0.134, 0.060, 0.064, 0.097, 0.132
  - Gross investment: 0.244, 0.267, 0.261, 0.372, 0.318
  - Exports: 0.272, 0.201, 0.241, 0.235, 0.224
- Theoretical proposition: the smaller the size of the government—measured as government expenditures in percent of GDP—the higher the import-to-GDP ratio, because government expenditure has a smaller import content than private consumption; a shift from government to private sector demand increases import demand.
- Empirical scatterplots for G-7 annual data 1990–2004 (Figure 1) using AMECO (general government) and IFS (central government) data:
  - Figure 1 shows a negative relationship between government expenditure ratio and import ratio for all G-7 countries except Japan.
  - Reported fitted lines (examples from panels):
    - General government: y = -0.90x + 0.44; y = -0.56x + 0.53; y = -1.04x + 0.84; y = 0.17x + 0.03 (t-statistics shown beneath coefficients).
    - Central government: y = -1.34x + 0.33; y = 0.42x + 0.02; y = -1.83x + 0.72; y = -0.86x + 0.43 (t-statistics shown).
  - With the exception of Japan, coefficients are significant at the 1 percent level.
  - Japan’s differing behavior may relate to its decade-long stagnation.
- Broader cross-country correlations (1990–2004, Table 2) indicate similar negative relationships between government expenditure ratio and import-to-GDP ratio across many countries:
  - Selected correlations (significance as reported):
    - Canada: -0.88***
    - Germany: -0.57***
    - United Kingdom: -0.59***
    - Italy: -0.94***
    - Japan: 0.42*
    - United States: -0.84***
    - Austria: -0.81***
    - Belgium: -0.89***
    - Sweden: -0.96***
    - Switzerland: 0.06
    - Norway: 0.43*
    - New EU Member States examples: Cyprus -0.98***; Hungary -0.73***; Malta -0.68***; Slovak Republic -0.62***.
  - Notes: *** 1 percent, ** 5 percent, * 10 percent significance. Data limitations for some countries mean correlations must be interpreted cautiously.
- Correlations indicate a robust negative relationship but do not quantify the impact of changes in government expenditure on imports or the current account.

### III. THE MODEL SPECIFICATION
- Standard trade model: export and import volume equations relate M and X to domestic income (Y), foreign income (Y*), and relative prices (RP). Log-form equations given as (1) and (2).
  - Exports (log form): rpx y x x t γ2 γ1 γ0 (equation (1) as presented).
  - Imports (log form): rpm y m m t δ2 δ1 δ0 (equation (2) as presented).
- Rationale for disaggregation: prior literature often assumes common elasticities for private consumption and government expenditure; authors disaggregate domestic real income into private consumption (C), private investment (I), government expenditure (G), and exports (X) to allow divergent import elasticities (equation (3)).
- Extended import equations separate goods and services and allow distinct elasticities for ln C, ln I, ln G, and ln X:
  - Goods imports extended (equation (8)).
  - Service imports extended (equation (9)).
- Data: annual G-7 panel from 1970 through 2002 (with services series from 1977 through 2002 where needed). World merchandise trade proxies world demand for goods exports (yg*); world real GDP proxies world demand for services (ys*).
- Theoretical expectations:
  - Income elasticities close to 1 but may deviate empirically.
  - Sum of demand elasticities (consumption, investment, government, exports) should be equal or close to 1 in extended trade equations.
  - Relative price coefficients expected to be negative.

### IV. EMPIRICAL ANALYSIS AND RESULTS
- Methodology:
  - Panel unit root tests: Levin, Lin, and Chu (LLC) and Breitung applied. Almost all variables are integrated of order one; cointegration techniques applied.
  - Panel cointegration tests: Pedroni (Phillips-Perron-type) and Kao (DF/ADF-like) reject the null of no cointegration in all cases for the trade volume equations (Table 4).
  - Estimation technique: pooled mean group (PMG) estimator (Pesaran and others (1999)); country-by-country Johansen/VEC and PMG/MG comparisons used; Hausman tests applied for slope homogeneity.
- Panel unit root test results (Table 3 highlights):
  - LLC and Breitung p-values reported for series (examples):
    - Relative price of exported goods rpxg: LLC 0.0579; Breitung 0.0584
    - Relative price of imported goods rpmg: LLC 0.8771; Breitung 0.4112
    - World real GDP ys*: LLC 0.0001; Breitung 0.9760
    - Private consumption c: LLC 0.0000; Breitung 0.8933
    - Government consumption g: LLC 0.0001; Breitung 0.2520
  - Conclusion: Breitung supports unit roots for most series; cointegration appropriate.
- Panel cointegration tests (Table 4): Kao DF-roh, DF-t, DF-rho*, DF-t*, ADF and Pedroni PC1/PC2 statistics reported with p-values; null of no cointegration rejected at conventional significance levels for goods exports, service exports, goods imports, service imports, extended goods imports, extended service imports.
- Conventional trade volume elasticities (Table 5, PMG estimates):
  - Goods export price elasticity: -0.849** (t = -8.647)
  - Goods export income elasticity: 0.906** (t = 36.395)
  - Service export price elasticity: -0.726** (t = -3.500)
  - Service export income elasticity: 1.018** (t = 3.572)
  - Goods import price elasticity: -0.313** (t = -3.076)
  - Goods import income elasticity: 1.953** (t = 9.896)
  - Service import price elasticity: -1.263** (t = -15.921)
  - Service import income elasticity: 1.316** (t = 56.190)
  - Joint Hausman tests reported (examples: 0.66, 0.89, 0.31, 0.94) indicate null of slope homogeneity not rejected for estimations.
- Extended import volume estimation results (Table 6, PMG estimates):
  - Goods imports (extended):
    - Price elasticity: -0.665** (-5.015)
    - Private consumption (ln C): 1.102** (3.481)
    - Government expenditure (ln G): 0.392* (1.762)
    - Private sector investments (ln I): 0.427** (5.152)
    - Exports (ln X): 0.435** (4.156)
    - Joint Hausman test: 0.12
  - Service imports (extended):
    - Price elasticity: -1.592** (-6.747)
    - Private consumption (ln C): 1.433** (1.916)
    - Government expenditure (ln G): 0.491** (2.485)
    - Private sector investments (ln I): 0.030 (0.076) [not significant]
    - Exports (ln X): 0.503** (1.972)
    - Joint Hausman test: 0.22
  - Significance notation: * 5 percent, ** 1 percent. t-statistics in parentheses.
- Key empirical findings:
  - The composition of demand matters: elasticities differ across private consumption, private investment, government expenditure, and exports.
  - Government expenditure has the smallest import elasticity among demand components but is positive and significant.
  - A lasting increase in government expenditure of 1 percent leads to:
    - an increase of goods imports of 0.4 percent (0.392* reported).
    - an increase of service imports of 0.5 percent (0.491** reported).
  - Therefore, ceteris paribus, an increase in government expenditure leads to a deterioration of the trade account.
- Caveats and indirect effects:
  - The ceteris paribus interpretation is problematic because fiscal changes may crowd out or crowd in private demand components:
    - Prior studies indicate government expenditure may crowd out private sector investment while increasing private consumption.
    - If government spending crowds out private investment but raises private consumption, effects on import volumes are ambiguous and depend on relative elasticities and magnitudes of change.
  - For services, results are more predictable: increases in government expenditure and related rises in private consumption increase service imports; private investment changes do not materially affect service imports.
  - Because the goods account is larger than the service account, goods-account effects likely dominate overall trade-account outcomes.

### V. SUMMARY AND CONCLUSION
- Main conclusions:
  - Fiscal policy matters for the trade account through both direct and indirect channels.
  - Disaggregating demand shows that government and private demand components have different import elasticities; previous literature that assumed common elasticities missed this.
  - Empirical PMG panel estimation for G-7 (1970–2002) finds that an increase in government expenditures by 1 percent leads to:
    - an increase in goods imports of about 0.4 percent.
    - an increase in service imports of almost 0.5 percent.
  - Implication: ceteris paribus, an increase in government expenditure would lead to a deterioration of the trade account.
- Policy-relevant interpretation:
  - The direct effect of higher government expenditure is to raise import demand and worsen the trade account, but indirect effects through private consumption and investment responses create ambiguity.
  - The overall impact of fiscal expansions or contractions on the current account depends on (a) the relative import elasticities of demand components and (b) how private consumption and investment respond to fiscal changes.
  - Given the empirical uncertainty about private-sector responses to fiscal policy, policymakers should be cautious in assuming a simple one-to-one relationship between changes in government spending and trade-account outcomes.

*Source: conclusions. (content unit _wp06147 - conclusions.)*

### conclusions if an increase (decrease) in government expenditure was to crowd out (crowd in)

### _wp06147 - conclusions if an increase (decrease) in government expenditure was to crowd out (crowd in)

### Ambiguity of fiscal impact on import demand
- Findings are ambiguous: an increase (decrease) in government expenditure could crowd out (crowd in) private demand; "If this crowding-in/out effect is strong enough, an increase in government expenditures could bring about the opposite result."
- The paper's results align with existing literature that finds divergent effects of fiscal policy on the trade account.
- Specific empirical findings cited:
  - Blanchard and Perotti (2002): fiscal expansion has a positive impact on consumption and a negative impact on investment.
  - Fatás and Mihov (2001): consumption increases in response to a positive expenditure shock, while investment is not affected significantly.
  - Karras (1994): private consumption and government spending are complementary; private consumption decreases as government expenditures are cut.
- Empirical context: "In the case of the G-7 countries, service imports are less than one-third of the size of goods imports."

### Compositional effect of government expenditure on aggregate demand
- Core result: higher government expenditures, ceteris paribus, lead to higher imports because the government "consumes more from abroad in line with the import content of government consumption."
- However, when accounting for the compositional effect of fiscal policy on overall demand—i.e., how private demand reacts—the opposite conclusion can also be derived.
- The ambiguity arises in part because private and public demand have differing trade elasticities.

### Implications and recommended further research
- The study reveals a difference between the trade elasticities of private and public demand.
- Recommended next steps: "Further research could determine the overall impact (i.e. the direct impact of a change in expenditure and the indirect impact through the reaction of private demand) that a change in government expenditure could have on the trade account of a particular country."
- Suggested methodology: "For this purpose, a country-specific analysis of the link between fiscal policy measures and private demand would be appropriate."

### Evidence from the literature (selected)
- The paper notes divergent findings in the literature on fiscal policy and the trade/current-account relationship, citing:
  - Erceg, Guerrieri, and Gust (2005); Lane and Perotti (1998); Baxter (1995) — divergent effects of fiscal policies on the trade account.
  - Bernheim (1988); Bussière, Fratzscher, and Müller (2004); Normandin (1999); Piersanti (2000); Enders and Lee (1990); Dewald and Ulan (1990); Kim and Roubini (2004) — differing results on the relation between fiscal deficit and current account deficit (some positive, some negative, some no significant relation).

### Data, variables, and sample (appendix summary)
- Sample: annual data for the G-7 countries (Japan, the United States, Canada, the United Kingdom, France, Italy, and Germany).
- Coverage period for trade-equation estimation: 1970 through 2002.
- Data sources: IMF’s IFS, OECD’s main economic indicators (MEI), IMF’s Direction of Trade Statistics (DOTS).
- Key variables and exact data definitions (preserved terminology and line numbers):
  - XG — Goods export volumes: "Export volumes (IFS line 72) ... using the 1995 average for merchandise exports in US$ (IFS line 78aa) ... deflating by PC."
  - XS — Service export volumes: "Service credits in US$ (IFS line 78ad) ... deflating by PC."
  - MG — Domestic goods import volumes: "import volume FOB series (IFS line 73) ... merchandise exports in US$ (IFS line 78ab) ... using the 1995 average for the US$ exchange rate (r)."
  - MS — Domestic service import volumes: "Service debits in US$ (IFS line 78ae) ... deflating by PCW after converting PCW into domestic currency terms using EFEX."
  - YG* — World income relevant for goods export demand: "total world exports in US$ at current prices (IFS line 70), deflated using WPXG."
  - YS* — World income relevant for service export demand: "Total OECD GDP at constant market prices in US$."
  - Y — Domestic real GDP: "IFS line 99b and deflated by PY."
  - C — real private consumption: "IFS line 96f and deflated by CP."
  - I — Real private sector investment: "IFS line 93i plus IFS line 93e and deflated by PY."
  - G — real government expenditure: "IFS line 91f and deflated by PY."
  - X — real exports: "IFS line 90c and deflated by PY."
  - PC — domestic consumer price index in domestic currency: "IFS line 64."
  - PCW — world consumer price index: "MEI of the OECD."
  - PXG — domestic export prices: "Export prices index (IFS line 76)... PD as the domestic prices index ... IFS line 63."
  - WPXG — world export prices in US$: "unit value of world exports in US$ (IFS line 74)."
  - PY — domestic GDP deflator: "IFS line 99bi."
  - r — nominal US$ exchange rate: "IFS line rf."
  - EFEX — nominal effective exchange rate: "Calculated from the exchange rates (r) and the bilateral trade weights (exports plus imports (lines 70 and 71 of the direction of trade statistics)) ... of the G-7 countries and their 39 largest trading partners."
  - RPXG — relative price for goods exports: "(WPXG*r)/PXG."
  - RMPG — relative price for goods imports: "(WPXG*r)/PD."
  - RPS — relative price for service exports and service imports: "PCW/ (PC*EFEX)."

*Italic source: _wp06147 - conclusions if an increase (decrease) in government expenditure was to crowd out (crowd in)_*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06147.pdf_