## _wp12286 - Appendix 1 of the April 2012 IMF Fiscal Monitor

## Source details

**Canonical URL:** [_wp12286 - Appendix 1 of the April 2012 IMF Fiscal Monitor](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12286.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12286.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12286.pdf.json)

---

### Main objective and contribution
- Investigates how the effects of discretionary fiscal policy on output vary depending on whether the economy is in an expansion or a downturn, where expansions and downturns are defined by the sign of the output gap (positive and negative, respectively).
- Develops a dataset of quarterly government expenditure and revenue for six of the G7 economies (Canada, France, Germany, Japan, the United Kingdom and the United States; Italy excluded) going back to the 1970s for most countries.
- Uses a nonlinear threshold vector autoregressive model (TVAR) that separates observations into regimes based on a threshold variable (the output gap). Within each regime the model is linear, but regimes may switch after a fiscal shock depending on the level of the output gap.

### Key methodological features
- Threshold variable: output gap (sign determines expansion vs downturn).
- Model: threshold vector autoregressive model (TVAR) allowing regime-dependent slopes for government spending and revenue.
- Country-by-country estimation with threshold level chosen endogenously to maximize model fit.
- Structural identification via Blanchard and Perotti (2002) procedure; revenue elasticities constructed following OECD calculations.
- Impulse responses via generalized impulse response function (GIRF) methodology; shocks of specific size implemented (2 percent shocks referenced); forecast horizon up to about 15 quarters.
- Confidence bands constructed using parametric bootstrap with 500 stochastic simulations; 1 standard deviation (68 percent) confidence bands reported in figures.
- Robustness checks include using output growth as threshold variable and Cholesky decomposition; spending results remain robust.

### Principal empirical findings (state-dependence and magnitudes)
- State-dependence:
  - For the average of the G7 economies, government spending and revenue multipliers tend to be larger in downturns than in expansions.
  - Linear SVAR multipliers lie between the positive- and negative-regime multipliers; linear models thus tend to understate effects during downturns and overstate effects during expansions.
- Reported ranges and central reported values:
  - First-year cumulative spending multipliers reported in the literature range from 1.6 to 2.6 for expenditure shocks, and from 0.2 to 0.4 for tax shocks.
  - First-year cumulative multipliers for consolidations that began in expansions range from 0.3 to 1.6 for expenditure shocks, and -0.3 to 0.2 for tax shocks.
  - Second-year cumulative multipliers have similar sizes to 1-year multipliers (a large part of the impact materializes within 4 quarters).
- Country-specific summaries:
  - Germany, Japan, and the United States: spending impact multipliers are statistically significant and sizeable; spending shocks have a significantly larger effect on output when the output gap is negative than when it is positive.
  - United Kingdom: spending multipliers are small under both positive and negative output gaps.
  - Revenue multipliers: generally less conclusive but statistically significant for Canada, France, Germany, and Japan.
  - Germany: revenue multipliers are slightly higher in “good times” than in “bad times.”
  - Canada and Japan: revenue measures work as a countercyclical tool only when the output gap is negative.
- Average G7 (excluding Italy) summary metric:
  - Assuming two thirds of adjustment comes from spending measures, a weighted average of spending and revenue multipliers in downturns yields an overall fiscal multiplier of about 1.

### Selected numeric examples reported in figures
- Average G7 standardized four-quarter cumulative multipliers (selected markers in source figure):
  - Positive spending shock (positive output gap): 0.72
  - Negative spending shock (negative output gap): -0.78
  - Negative revenue shock (positive output gap): -0.04
  - Positive revenue shock (negative output gap): 0.03
  - Other numeric markers shown in graphical depictions include: 0.79, -0.79, 0.29, -0.29, 1.22, -1.34, 0.35, -0.40
- Examples of country figure values (four- and eight-quarter cumulative multipliers as shown in source figures):
  - Canada: example bars include -0.9, -1.1 (after four quarters for some bars).
  - France: example bars include 0.0, 0.2, -0.1, -0.2.
  - Germany: example bars include 0.2, 1.0, 0.7, 0.5 and larger downturn values 1.4, 2.0.
  - Japan: example bars include 1.4, 2.0 and some small negative values.
  - United Kingdom: example bars include 0.0, 0.2, 0.4.
  - United States: spending multipliers significantly above one during downturns reported.
- Impulse experiments and axis scales in figures:
  - Fiscal shock labels such as "Japan - 2 % Positive Spending Shock", "Canada - 2 % Negative Revenue Shock", and corresponding -2 % contraction variants are used.
  - Figure axis tick ranges include sequences like -1.6 to 0.0 and panel ranges such as -3.0, -1.0, 1.0, 3.0; -4.0 to 6.0; -6.0 to 4.0; and -0.8 to 0.8.
  - Confidence bands plotted at 68 percent.

### Data and sample details
- Countries: Canada, France, Germany, Japan, United Kingdom, United States. Italy excluded for lack of comparable quarterly fiscal data.
- Sample periods (selected exact entries):
  - Canada: 1966Q1–2011Q2.
  - France: 1970Q4–2010Q4.
  - Germany: 1975Q3–2009Q4.
  - Japan: 1970Q1–2011Q2.
  - United Kingdom: 1970Q1–2011Q2.
  - United States: 1965Q2–2011Q2.
- Fiscal data definitions:
  - Net revenue = general government revenues minus net transfers.
  - Government spending = general government investment and consumption (excluding transfers and subsidies).
  - Fiscal data cover general government and are deflated with the GDP deflator.
- Output gap data mainly from OECD; for Germany the HP filter is used (λ =1600).
- VAR specification: three variables—real GDP, real net revenue, real net expenditure—following Blanchard and Perotti (2002).

### Robustness, sensitivity, and identification notes
- Threshold selection: Tsay (1998) multivariate threshold approach used to test for threshold behavior and select threshold lag d (tested 1 to 8 lags); estimated threshold values are negative for most countries and significant at:
  - 1 percent for Canada, Germany, Japan and the United States;
  - 5 percent for the United Kingdom;
  - 10 percent for France.
- Robustness checks: estimations using output growth as threshold variable yield qualitatively similar results except for France; Cholesky decomposition used as a check.
- GIRF-specific considerations noted: trend GDP treated as given within forecast horizon; output gap persistence and lag structure can affect early-quarter IRF sensitivity to shock sign.

### Policy implications and scenarios
- Timing and pace of consolidation:
  - When the output gap is initially negative, fiscal adjustment implemented gradually has a smaller negative impact on growth (cumulative over two and one-half years) than an up-front consolidation of the same overall size.
  - When the output gap is initially positive, a front-loaded tightening has a smaller cumulative impact on growth.
  - Gradual consolidation increases the probability of moving from a negative-output-gap regime into a positive-output-gap regime while tightening is in progress, reducing short-term output costs.
- Design and sequencing:
  - Heterogeneity across countries implies optimal sequencing and composition of fiscal measures should be tailored to country-specific multiplier estimates and cyclical positions.
  - High pre-existing tax pressure suggests bulk of adjustment, when feasible, should focus on the expenditure side, though revenue increases may be unavoidable for large adjustments.
- Practical guidance:
  - If financing allows and the economy starts with a negative output gap, consider gradual (phased) fiscal consolidation to reduce short-term negative growth effects.
  - In positive-output-gap periods, more front-loaded consolidation can be less harmful to cumulative growth.
- Suggested research extensions:
  - Investigate interaction between fiscal multipliers and monetary policy, particularly near the zero lower bound.
  - Analyze multiplier effects of different revenue and expenditure components by state of the economy.
  - Extend sample to other advanced and emerging economies.

### Caveats and limitations
- Substantial cross-country heterogeneity in multipliers; country-by-country estimates are important.
- Revenue multipliers results are generally less conclusive than for spending multipliers.
- Model limitations:
  - VAR includes only three variables and does not explicitly model interactions with monetary policy, public debt, or financial market responses.
  - Analysis may overestimate multipliers in high debt countries.
  - Blanchard and Perotti (2002) identification may not fully capture exogenous policy changes and can be affected by asset and commodity price movements; corrections applied remove largest measurement errors but not all.
- Data limitations:
  - Data interpolation and quality issues for Japan and France; true quarterly fiscal data for France available only since the 1990s.
  - IMF (2010) action-based (narrative) measures are annual and not directly comparable to the quarterly analysis.
- Additional notes:
  - Anticipation effects are not modeled; gradual adjustment could induce different pre-announcement dynamics.
  - Potential long-term costs of sharp consolidation (employment, human capital, financial markets) are not captured.
  - Sharp up-front adjustment may affect market confidence and bond spreads; source notes bond spreads in the contemporaneous sovereign debt crisis appear largely driven by GDP growth prospects.

*Source: Appendix 1 of the April 2012 IMF Fiscal Monitor (IMF staff estimates and analysis).*

### Appendix 1 of the April 2012 IMF Fiscal Monitor.

### _wp12286 - Appendix 1 of the April 2012 IMF Fiscal Monitor

### Main objective and contribution
- Investigates how the effects of discretionary fiscal policy on output vary depending on whether the economy is in an expansion or a downturn, where expansions and downturns are defined by the sign of the output gap (positive and negative, respectively).
- Develops a dataset of quarterly data on government expenditure and revenue for six of the G7 economies (excluding Italy) going back to the 1970s.
- Uses a nonlinear threshold vector autoregressive model (TVAR) that separates observations into regimes based on a threshold variable (the output gap). Within each regime the model is linear, but regimes may switch after a fiscal shock depending on the level of the output gap.

### Key methodological features
- Threshold variable: output gap (sign determines expansion vs downturn).
- Model: threshold vector autoregressive model (TVAR) allowing regime-dependent slopes for government spending and revenue.
- Country-by-country estimation: explanatory variables (government spending and revenue) allowed to have differing regression slopes across regimes; threshold level chosen to maximize model fit.
- Rationale for output gap threshold: under a negative output gap excess capacities reduce crowding out of private investment after a government spending shock; a higher share of credit-constrained households and firms increases responsiveness to disposable income changes.

### Principal empirical findings
- For an average of G7 economies, government spending and revenue multipliers tend to be larger in downturns than in expansions.
- Multipliers differ significantly across countries, implying the need for country-specific assessment and policy design.
- Country-level summary (as reported):
  - Germany, Japan, and the United States: spending impact multipliers are statistically significant and sizeable; spending shocks have a significantly larger effect on output when the output gap is negative than when it is positive.
  - United Kingdom: spending multipliers are small under both positive and negative output gaps.
  - Revenue multipliers: results generally less conclusive, but impacts are more significant for Canada, France, Germany, and Japan.
  - Germany: revenue multipliers are slightly higher in “good times” than in “bad times,” consistent with less Ricardian behavior when market sentiment is positive.
  - Canada and Japan: revenue measures work as a countercyclical tool only when the output gap is negative.

### Literature context and summary statistics from reviews
- Baunsgaard and others (2012) review 37 studies (DSGE and VAR approaches); reported ranges and means:
  - Government spending multipliers range between 0 and 2.1, with a mean of 0.8 during the first year after fiscal measures are taken.
  - Government revenue multipliers range from about –1.5 to 1.4, with a mean of 0.3.
- Multipliers tend to be smaller in more open economies and in countries with larger automatic stabilizers.
- Few studies distinguish multipliers by state of the economy; notable prior country studies include Baum and Koester (2011) for Germany and Auerbach and Gorodnichenko (2012a) for the United States; panel approaches (Auerbach and Gorodnichenko, 2012b) may mask heterogeneities.

### Policy implications and scenarios
- Timing and pace of fiscal consolidation matter when the output gap is negative:
  - When the output gap is initially negative, fiscal adjustment implemented gradually has a smaller negative impact on growth (cumulative over two and one-half years) than does an up-front consolidation of the same overall size.
  - This suggests that, when feasible, a more gradual fiscal consolidation may be preferable to an approach that concentrates adjustment up-front.
- Heterogeneity across countries implies that optimal sequencing and composition of fiscal measures should be tailored to country-specific multiplier estimates and cyclical positions.

### Caveats and limitations (as discussed)
- Multipliers vary widely across countries; country-by-country heterogeneity is significant.
- Results for revenue multipliers are generally less conclusive than for spending multipliers.
- Identification and regime classification depend on the chosen threshold variable and model specification (TVAR); results reflect those methodological choices.

_Italic source: Appendix 1 of the April 2012 IMF Fiscal Monitor._

### 1.6 to 2.6 for expenditure shocks, and 0.2 to 0.4 for tax shocks. First-year cumulative

### _wp12286 - 1.6 to 2.6 for expenditure shocks, and 0.2 to 0.4 for tax shocks. First-year cumulative

### Key findings
- First-year cumulative spending multipliers reported in the literature range from 1.6 to 2.6 for expenditure shocks, and from 0.2 to 0.4 for tax shocks.
- First-year cumulative multipliers for consolidations that began in expansions range from 0.3 to 1.6 for expenditure shocks, and -0.3 to 0.2 for tax shocks.
- Second-year cumulative multipliers have similar sizes to 1-year multipliers, implying that a large part of the impact of fiscal shocks on output materializes within 4 quarters.
- Across the six G7 countries analyzed (Canada, France, Germany, Japan, the United Kingdom and the United States):
  - Government spending shocks generally have a larger effect on output when the output gap is negative (downturns), with Canada as the only exception.
  - Spending multipliers are sizeable and statistically significant on impact for Germany, Japan, and the United States.
  - First-year revenue multipliers are small on average (well below 0.5) and statistically significant for Canada, France, Germany, and Japan.
  - In Germany, revenue multipliers are slightly higher in “good times” than in “bad times”.
  - In Canada and Japan, revenue measures act as a countercyclical tool mainly when the output gap is negative.
- For the average of the G7 economies (excluding Italy), a weighted average of spending and revenue multipliers in downturns—assuming two thirds of adjustment comes from spending measures—yields an overall fiscal multiplier of about 1.
- Linear (standard) SVAR multipliers lie between the positive- and negative-regime multipliers; linear models thus tend to understate effects during downturns and overstate effects during expansions.

### Data and sample
- Countries in sample: Canada, France, Germany, Japan, the United Kingdom and the United States.
- Quarterly datasets constructed for most countries since at least the 1970s; Italy excluded due to lack of comparable quarterly fiscal data.
- Data sources include OECD Economic Outlook, IMF’s International Financial Statistics, Eurostat and national account data.
- Fiscal data cover the general government. Net revenue = general government revenues minus net transfers. Government spending = general government investment and consumption (excluding transfers and subsidies). All series deflated with the GDP deflator.
- Output gap data obtained from the OECD for most countries; for Germany the HP filter is used.
- VAR specification: three variables—real GDP, real net revenue, real net expenditure—following Blanchard and Perotti (2002).

### Methodology
- Nonlinear time-series analysis using a threshold VAR following Baum and Koester (2011):
  - Threshold value determined endogenously to maximize fit in both regimes; model focuses on two regimes (positive and negative output gap).
  - Tsay (1998) multivariate threshold approach used to test for threshold behavior and select threshold lag d (tested 1 to 8 lags).
- Identification:
  - Structural identification via Blanchard and Perotti (2002) procedure, accounting for automatic stabilizers.
  - Revenue elasticities with respect to GDP obtained following OECD calculations (Girouard and André, 2005); quarterly weighted elasticities constructed from shares of tax types and respective elasticities.
  - Net revenue and expenditure series corrected to remove changes not linked to fiscal policy decisions or cyclical adjustments where possible (e.g., large asset or commodity price movements).
- Impulse responses:
  - Generalized impulse response function (GIRF) methodology (Koop, 1996; Koop, Pesaran, and Potter, 1996) employed to account for nonlinearity and possible regime switches after shocks.
  - Shocks of specific size implemented; 2 percent shocks referenced as reasonable size used.
  - Forecast horizon up to about 15 quarters.
  - Trend GDP treated as given within the forecast horizon (trend unaltered by fiscal shocks for several quarters).
  - Confidence bands constructed using standard parametric bootstrap following Luetkepohl (2000) with 500 stochastic simulations; 1 standard deviation confidence bands taken from distribution of resulting IRFs.
- Robustness checks:
  - Estimations also performed using output growth as threshold variable (results qualitatively similar except for France).
  - Cholesky decomposition applied as a robustness check; spending results remain robust.

### Country-by-country and graphical results (selected exact values and observations)
- Table 3 descriptive statistics (threshold estimation): estimated threshold values are negative for most countries and significant at:
  - 1 percent for Canada, Germany, Japan and the United States;
  - 5 percent for the United Kingdom;
  - 10 percent for France.
- Figures 2 and 3 (four- and eight-quarter cumulative multipliers) report country-specific numeric multipliers for positive vs. negative output gap regimes. Selected reported values shown in the source figures include (values are four- and eight-quarter cumulative multipliers presented for fiscal expansion and contraction):
  - Canada: example values in figures include negative and small magnitudes (e.g., after four quarters: -0.9, -1.1 shown for some bars).
  - France: sample figure values include 0.0, 0.2, -0.1, -0.2 in various bars.
  - Germany: examples include 0.2, 1.0, 0.7, 0.5 and larger values in downturns (e.g., 1.4, 2.0).
  - Japan: examples include 1.4, 2.0 and negative small values in some cases.
  - United Kingdom: examples include small positive and negative values (e.g., 0.0, 0.2, 0.4).
  - United States: spending multipliers significantly above one during downturns are reported.
- Note on significance: striped bars in figures indicate measures for which no significant multiplier is found at the time the fiscal shock is implemented.
- Average G7 figure (Figure 4) reports standardized cumulative multipliers over four quarters with selected numeric markers in the source figure:
  - Positive spending shock (positive output gap): 0.72
  - Negative spending shock (negative output gap): -0.78
  - Negative revenue shock (positive output gap): -0.04
  - Positive revenue shock (negative output gap): 0.03
  - Linear model averages and negative output gap bars also display numeric markers such as 0.79, -0.79, 0.29, -0.29, 1.22, -1.34, 0.35, -0.40 in the graphical depiction.

### Interpretation and comparisons with other literature
- Results confirm state dependency of fiscal multipliers: spending multipliers are significantly larger in downturns than in expansions.
- Spending multipliers in the United States found to be significantly above one during downturns.
- Revenue multipliers are generally smaller than spending multipliers; first-year revenue multipliers typically well below 0.5 on average.
- Comparisons:
  - Findings broadly align with Auerbach and Gorodnichenko (2012a) and Batini, Callegari, and Melina (2012) regarding state-dependent multipliers.
  - Results differ from narrative-based studies (Romer and Romer, 2010; Cloyne, 2011) which find large revenue multipliers for the United States and the United Kingdom; potential upward biases in those narrative estimates are noted in subsequent work (Favero and Giavazzi, 2012).
  - Linear SVAR multipliers lie between the expansion and downturn regime multipliers and match averages identified in prior literature (Baunsgaard and others, 2012).

### Caveats and limitations
- Model includes only three variables and does not explicitly model interactions with monetary policy and public debt.
  - Previous work (Auerbach and Gorodnichenko (2012b); Ilzetzki, Mendoza, and Vegh (2010)) suggests government debt, level of development, exchange rate flexibility, and openness can influence fiscal multiplier size.
  - Analysis may overestimate multipliers in high debt countries.
- Data limitations:
  - Data interpolation and quality issues for Japan and France; true quarterly data for France available only since the 1990s.
  - IMF (2010) action-based (narrative) measures are annual and thus not directly applicable to the quarterly analysis.
- Identification and measurement issues:
  - Blanchard and Perotti (2002) BP approach may not fully capture exogenous policy changes and can be affected by asset and commodity price movements, potentially biasing estimates.
  - Corrections applied to net revenue and expenditure series remove the largest measurement errors but not all.
- GIRF-specific considerations:
  - Output gap persistence and treatment of trend GDP (trend held given within forecast horizon) reduce differences between positive and negative shocks.
  - Output gap entering VAR with lags can make early-quarter IRFs insensitive to shock sign in some specifications.
- Additional caveats on policy implications:
  - Anticipation effects are not modeled; gradual adjustment could induce different pre-announcement dynamics.
  - Model does not capture broader potential long-term costs of sharp consolidation (employment, human capital, financial markets).
  - Sharp up-front adjustment may change market confidence and bond spreads; the source notes bond spreads in the current sovereign debt crisis appear largely driven by GDP growth prospects.

### Policy implications and scenarios
- Up-front versus gradual implementation:
  - When the output gap is negative initially, gradual fiscal adjustment (spread evenly over two years) hurts growth less in the short term than a front-loaded improvement; this is illustrated for the G7 average (excluding Italy) in Figure 5.
  - Conversely, when the output gap is initially positive, a front-loaded tightening has a smaller cumulative impact on growth.
  - Nonlinearity and regime-switching imply that a gradual consolidation increases the probability of moving from a negative-output-gap regime into a positive-output-gap regime while tightening is in progress, reducing short-term output costs.
- Design considerations:
  - Given heterogeneity of multipliers across and within countries, fiscal policy design should be tailored country-by-country.
  - High pre-existing tax pressure in some countries suggests the bulk of adjustment, when feasible, should focus on the expenditure side despite potential inevitability of some revenue increases for large targeted adjustments.
  - Long-term effects and efficiency of specific tax and expenditure changes depend on preexisting levels and structures of public finances.
- Practical guidance:
  - If financing allows and the economy starts with a negative output gap, consider gradual (phased) fiscal consolidation to reduce short-term negative growth effects.
  - In positive output gap periods, more front-loaded consolidation can be less harmful to cumulative growth.

*Source: IMF staff estimates and analysis in the referenced chapter.*

### Annex 3 of IMF (2012) also finds that provided that hysteresis effects are significant and multipliers are

### _wp12286 - Annex 3 of IMF (2012) also finds that provided that hysteresis effects are significant and multipliers are 

### Major findings on fiscal multipliers and state-dependence
- Annex 3 finds that provided that hysteresis effects are significant and multipliers are asymmetric across the business cycle, delaying consolidation can generate permanent output gains.
- The impact of fiscal policy on economic activity varies with the business cycle and is nonlinear.
- Fiscal multipliers for the six economies analyzed are on average larger in times of negative output gaps than when the output gap is positive.
- Spending shocks tend to have a larger effect on output when the output gap is negative, particularly in those countries where spending impact multipliers are statistically significant and sizeable:
  - Germany
  - Japan
  - United States
- Revenue multipliers show more heterogeneous results:
  - For Canada, France, Germany, and Japan the impact is statistically significant.
  - In Germany, revenue multipliers are slightly higher in “good times” than in “bad times”.
  - In Canada and Japan, revenue measures work as a countercyclical tool only when the output gap is negative.
- The heterogeneity of multipliers across countries implies the need for tailored, country-by-country assessment of fiscal policy effects.

### Quantitative results, shocks, and confidence reporting
- Figures illustrate cumulative impact on output from discretionary fiscal spending shocks for G-7 economies, with axis tick labels including:
  - Vertical scale examples: -1.6, -1.4, -1.2, -1.0, -0.8, -0.6, -0.4, -0.2, 0.0
  - Horizontal axis: 1 2 3 4 5 6 7 8 9 10 11 12 13 (quarters)
- Fiscal stimulus/contraction impulse-response exercises use shocks described in the figures as:
  - Japan - 2 % Positive Spending Shock
  - Japan - 2 % Negative Revenue Shock
  - Canada - 2 % Positive Spending Shock
  - Canada - 2 % Negative Revenue Shock
  - France - 2 % Positive Spending Shock
  - France - 2 % Negative Revenue Shock
  - Germany - 2 % Positive Spending Shock
  - Germany - 2 % Negative Revenue Shock
  - UK - 2 % Positive Spending Shock
  - UK - 2 % Negative Revenue Shock
  - US - 2 % Positive Spending Shock
  - US - 2 % Negative Revenue Shock
  - And corresponding - 2 % contraction variants (negative spending, positive revenue) for fiscal contraction figures.
- Confidence bands reported as 68 percent in Figures B1 and B2: "Cumulative Global Impulse Response Functions and Confidence Bands (68 percent)".
- Figure axis ranges cited across panels include examples:
  - -3.0, -1.0, 1.0, 3.0
  - -4.0, -2.0, 0.0, 2.0, 4.0, 6.0
  - -6.0, -4.0, -2.0, 0.0, 2.0, 4.0
  - -0.8, -0.6, -0.4, -0.2, 0.0, 0.2, 0.4, 0.6, 0.8

### Data, samples, and methodological notes
- Country sample periods and data definitions (selected exact entries):
  - Canada: 1966Q1–2011Q2. Fiscal data: Taxes on income+contributions to social insurance plan+taxes on production and income+(other current transfers from persons-current transfers paid). Expenditure: Outlay-Interest paid on public debt-current transfers. Sources: CANSIM National Income and Expenditure Accounts. Fiscal Data on general government and accrual basis in national currency. Data were seasonally adjusted with Tramo Seats in EViews. GDP and output gap from OECD; GDP deflator base 2002.
  - France: 1970Q4–2010Q4. Fiscal data: Tax revenue+social security contributions+other current transfers-social transfers-subsidies. Total expenditure-interest payments-transfers-subsidies. Sources: IFS (until 1990) and Eurostat (since 1991). Data seasonally adjusted using Tramo Seats in EViews. GDP and output gap (OECD).
  - Germany: 1975Q3–2009Q4. Fiscal data per Baum and Koester (2011). Real GDP and deflator (2000=100) are seasonally adjusted by BV 4.1 procedure. Output gap calculated with HP filter (λ =1600).
  - Japan: 1970Q1–2011Q2. Fiscal data from OECD. Data are on a general government and accrual basis and already seasonally adjusted. GDP and output gap (OECD); GDP deflator base 2000.
  - United Kingdom: 1970Q1–2011Q2. Fiscal data from OECD; GDP and output gap (OECD); GDP deflator base 2000.
  - United States: 1965Q2–2011Q2. Fiscal data: Current tax receipts+ Contributions for government social insurance+ Current transfer receipts- current transfers-subsidies. Consumption expenditures+ government investment. Sources: Bureau of Economic Analysis, OECD. Real GDP (chained 2005 dollars), Implicit GDP deflator (2005=100). Output gap (OECD).
- Threshold-variable sensitivity:
  - Appendix C compares multipliers estimated using output gap or GDP growth as the threshold variable (Table C1 presents numerous paired numeric entries by country and 4 quarters / 8 quarters horizons; regimes labeled Positive/Negative output gap and Positive/Negative GDP growth).
- Estimation approach:
  - Extends literature using non linear estimation techniques and a new dataset for six of the G7 economies.
  - Uses Threshold VAR techniques and generalized impulse response analysis; confidence bands obtained via authors' calculations.

### Policy implications and recommendations
- The state-dependence of multipliers implies that the timing and pace of fiscal consolidation matter:
  - When the output gap is negative at implementation, an up-front negative fiscal spending shock will have a larger short-term impact on output than a more gradual fiscal adjustment.
  - Given cross-country heterogeneity in multipliers, fiscal policy should be tailored to country-specific conditions.
- Suggested extensions and research directions:
  - Investigate interaction between fiscal multipliers and monetary policy, particularly near the zero lower bound for interest rates.
  - Analyze multiplier effects of different revenue and expenditure components and their dependence on the state of the economy.
  - Extend the country sample to other advanced and emerging economies to examine state dependency of multipliers more broadly.

*Source: Authors' calculations and country data as presented in the Annex of the IMF working paper excerpt.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12286.pdf_