## _wp12150

## Source details

**Canonical URL:** [_wp12150](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12150.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12150.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12150.pdf.json)

---

### Theoretical benchmark: private consumption and multipliers
- Private consumption in modern business cycle models (neoclassical and new Keynesian) is governed by intertemporal optimization and generally falls in response to an increase in government spending.
- Consequence: output multipliers are considerably smaller than predicted by traditional Keynesian analysis, which predicts a positive consumption response.
- Baxter and King (1993): impact multipliers on output hardly ever exceed unity across various specifications.
- Extensions often remain within standard business cycle frameworks and frequently abstract from the exchange rate regime, government debt, and financial frictions.

### Theoretical benchmark: pegged vs. flexible exchange rates
- Mundell-Fleming textbook prediction:
  - Flexible exchange rates: fiscal expansion crowds out net exports one-for-one via exchange rate appreciation (assuming unchanged monetary stance).
  - Fixed exchange rates: fiscal policy is effective because monetary expansion offsets appreciation pressure.
- New Keynesian qualifications:
  - Fiscal stimulus can be more or less effective under a flexible exchange rate depending on monetary policy and medium-term debt consolidation (Corsetti, Kuester, and Müller (2011a)).
  - Most standard models predict higher government spending appreciates the real exchange rate.
  - Identified exceptions:
    - Kollmann (2010): government spending may depreciate the real exchange rate if spending shocks are very persistent and international financial markets are incomplete.
    - Ravn, Schmitt-Grohé, and Uribe (2012): with “deep habits” and imperfect competition, markups and prices may fall in the short run, causing the real exchange rate to depreciate.
  - Corsetti, Meier, and Müller (2011b): short-run exchange rate response depends on the medium-term debt consolidation regime; anticipated future spending cuts can prevent long-term real interest rate rises and allow real exchange rate depreciation.

### Theoretical benchmark: weak public finances
- Empirical puzzle: Giavazzi and Pagano (1990) — fiscal consolidations in Denmark and Ireland during the 1980s coincided with improving government balances and strong private consumption growth.
- Mechanisms and findings:
  - Bertola and Drazen (1993): correlation between private and public consumption changes can shift as debt approaches a critical level that triggers retrenchment.
  - Perotti (1999): with unconstrained and credit-constrained households and distortionary taxes:
    - In “good” times, fiscal balances and private consumption comove negatively.
    - In “bad” times (high public debt), the comovement can become positive.
    - Tax distortions with high initial debt can amplify negative wealth effects for unconstrained households.
  - Perotti (1999) empirical two-step: near-VAR to identify fiscal shocks; regression of consumption growth on shocks and interaction with “bad times” dummy. Result: consumption responds positively to spending increases and tax cuts in normal times; sign reverses in bad times.
  - Sutherland (1997) and Afonso and Jalles (2011): tax cuts may be contractionary or reductions in government consumption can raise private consumption when public debt is high.

### Theoretical benchmark: financial crises and the zero lower bound
- Financial crises limit credit access; models with a fraction of constrained households imply larger fiscal multipliers when the share of constrained households is higher (Galí et al. (2007); Bilbiie, Meier, and Müller (2008); Fernández-Villaverde (2010)).
- Corsetti et al. (2011a): credit constraints have a larger effect on multipliers under fixed exchange rates or in a monetary union.
- Zero lower bound implications:
  - Monetary policy constrained at zero can allow fiscal stimulus to be fully accommodated at unchanged (zero) policy rates, arresting deflationary dynamics (Eggertsson and Woodford (2003)).
  - Under zero lower bound conditions, fiscal multipliers on output can exceed two or three (Christiano et al. (2011); Hall (2009); Erceg and Lindé (2010); Woodford (2011)).
  - Caveat: sovereign risk premia from higher public deficits can raise funding costs and may alter multiplier magnitudes (Corsetti et al. (2012)).

### Empirical strategy overview and identification issues
- Two-stage estimation strategy (Perotti (1999)-style):
  1. Estimate a fiscal policy rule for each country to describe the statistical process of government spending and obtain spending shocks.
  2. Use estimated policy shocks as regressors to trace impacts on output, private consumption, investment, the trade balance, and the real effective exchange rate.
- Rationale:
  - Structural VAR approaches assume discretionary government spending is subject to decision/implementation lags and is unresponsive to contemporaneous conditions—plausible for government consumption and investment.
  - Two-step approach advantages:
    - Country-by-country estimation allows heterogeneity in policymaking.
    - Flexible specification to account for transmission differences across environments (peg vs. flexible; sound vs. strained finances; normal times vs. financial crisis).
- Identification caveats:
  - Key assumption: no two-way contemporaneous interdependence between spending and its determinants (spending cannot respond to simultaneous output developments).
  - Use of lagged CLI (October of the previous year) to capture expectations at budget formulation.
  - Annual data attenuate anticipation problems relative to quarterly data.
  - Arellano-Bond tests for serial correlation on country-specific residuals: null of no autocorrelation cannot be rejected for any country.

### First-step: fiscal rule specification and data
- Policy variable: per capita government consumption expressed in logs.
  - Government consumption: 21.5 percent of GDP in the average country in the sample.
- Fiscal rule estimated for each country i (equation given in the source):
  g_t,i = φ_i + η_i trend_t + β_i,1 g_{t−1,i} + β_i,2 g_{t−2,i} + γ_i,1 y_{t−1,i} + γ_i,2 y_{t−2,i} + θ_i cli_{t−1,i} + δ_i b_{t−1,i} + ρ_i,1 peg_{t−1,i} + ρ_i,2 strain_{t,i} + ρ_i,3 crisis_{t−1,i} + ε_{t,i}. (1)
- Included regressors and features:
  - First and second lags of government spending (g_{t−1}, g_{t−2}).
  - First two lags of log per capita output (y_{t−1}, y_{t−2}).
  - Lagged composite leading indicator cli_{t−1} (normalized OECD CLI from October of the previous year).
  - Beginning-of-period debt stock b_{t−1} (share of GDP).
  - Trend and constant.
  - Lagged-information dummies: peg_{t−1}, strain_{t}, crisis_{t−1}.
  - Parameters stable over time within country but heterogeneous across countries.
  - ε_{t,i} is the additive shock term capturing unexpected discretionary policy changes.

### First-step: sample, dummies, and estimation properties
- Data frequency and sample:
  - Annual data; maximum coverage 1975 through 2008.
  - Requirement: at least 20 consecutive annual observations per country.
  - Final sample: 17 countries: Australia, Austria, Belgium, Canada, Denmark, Finland, France, Ireland, Italy, Japan, Netherlands, Norway, Portugal, Spain, Sweden, the United Kingdom, and the United States.
  - Final sample size for second-stage estimation: 444 country-year observations.
- Dummy definitions and thresholds:
  - Exchange rate regime: Ilzetzki et al. (2009) classification.
  - Financial crisis dates: Reinhart and Rogoff (2008) and Reinhart (2010).
  - Fiscal strain (weak public finances): government debt in excess of 100 percent of GDP or net government borrowing above 6 percent of GDP (lagged once); alternative robustness threshold used: debt >120 percent of GDP or deficit >7 percent of GDP.
- Properties of estimated shocks (first-step residuals, final sample):
  - No. of observations: 444
  - Mean: 0.04
  - Median: 0.00
  - Standard deviation: 1.02
  - Minimum: -3.57
  - Maximum: 5.16
  - Correlation with simple growth rate of government spending: 0.64
  - Five largest negative shocks:
    - Portugal, 1993: -3.57
    - Netherlands, 1984: -3.33
    - Netherland, 2005: -3.18
    - Norway, 1988: -2.97
    - Spain, 1988: -2.67
  - Five largest positive shocks:
    - Portugal, 1991: 2.60
    - Portugal, 2005: 2.68
    - Denmark, 1993: 2.85
    - Ireland, 1986: 3.83
    - Netherlands, 2006: 5.16
- Qualitative first-step findings:
  - Good fit of the empirical fiscal rule (high fit due to autoregressive terms); robustness confirmed when re-estimated in growth rates.
  - Government spending tends to respond negatively to the outstanding stock of public debt:
    - Coefficient on debt negative for all but three countries (Austria, Finland, and the United Kingdom).
    - Negative coefficient statistically significant for about half of the countries.
  - Financial crisis dummy estimates: for countries experiencing financial crises within the sample, the estimated crisis dummy coefficient is negative in nine out of thirteen countries — suggesting government spending tended to slow down rather than accelerate during those financial crises.

### Role of first-step shocks
- Estimated shocks (ε̂_{t,i}) are used as inputs for second-step panel regressions tracing dynamic effects of government spending shocks on output, components, international prices, and other macro variables.
- Validation: shocks show no serial correlation and plausible distributional properties, supporting causal second-step analysis.

### Unconditional effects: estimation approach
- Unconditional estimates omit all economic environment dummies.
- Two-step procedure with 1,000 draws from multivariate normal distribution for standard errors (mean equal to point estimate; variance-covariance matrix of regression coefficients used).
- Impulse responses normalized so that the initial increase in government spending is one percent of GDP (the point estimate of the response of government spending to the standardized shock tends to be around 0.9, reflecting a moderate attenuating bias in the second step caused by measurement error in the first step).

### Unconditional effects: key impulse-response findings
- Government spending:
  - Persistent increase (normalized to one percent of GDP initial increase).
- Output:
  - Sizeable increase by about 0.7 percentage points on impact.
- Consumption:
  - Does not respond on impact.
- Investment:
  - Private investment falls (clear-cut negative response).
- Net exports:
  - Decline in net exports following the positive spending shock.
- Inflation:
  - Drops on impact, then rises temporarily above baseline.
- Short-term interest rate:
  - Rises on impact and then moves in tandem with inflation.
- Real effective exchange rate:
  - Appreciates slightly on impact, but then depreciates over time.
- Multipliers and comparisons:
  - Output multipliers reported in the literature: range from 0.5 to 1.0.
  - Consumption multipliers reported in the literature: range from somewhat negative to 0.5.
  - Alternative multiplier computation (cumulative response divided by cumulative spending response) yields very similar results.
- Interpretation:
  - Unconditional results align with the literature and serve as benchmark.
  - Negative investment and worsening trade balance are consistent with conventional findings.
  - Short-term decline in inflation on impact is surprising but documented elsewhere.
  - Unconditional estimates may conceal large heterogeneity across economic environments, motivating conditional analysis.

### Sensitivity analysis: alternative crisis and weak-finance definitions
- Narrower crisis dummy (only the “big five” banking crises plus the global financial crisis starting in 2007):
  - Government spending appears less persistent.
  - Impact response for output and consumption remains close to 2.
  - Activity reverts to trend more quickly.
  - Inflation and real exchange rate responses stronger; sharper rise in short-term interest rate.
  - Multipliers conditional on high debt unaffected.
  - Under the peg, differences in impact responses across regimes become larger.
  - The puzzling delayed rise in consumption disappears.
- Stricter weak-public-finances thresholds (debt >120 percent of GDP and/or lagged net borrowing >7 percent of GDP):
  - Conditional on financial crisis and weak public finances: not substantially affected.
  - Under a peg, point estimates of impact responses for output and investment are lower relative to flexible exchange rates (reversal of prior main result).
  - External variables show similar or stronger responses: real exchange rate appreciates on impact and trade deficit worsens.

### Sensitivity analysis: specification and sample checks
- Specification in differences (vs. levels):
  - Results robust to differences specification.
  - Unexpected spending increases remain considerably more effective during financial crises than in normal times.
  - Output and consumption multipliers remain around 2.
  - Investment rises over time; net exports deteriorate; real exchange rate depreciates.
  - Conditional results on weak public finances and exchange rate peg confirmed; under a peg the output multiplier aligns more with received wisdom.
- Contemporaneous crisis dummy (vs. lagged):
  - Including contemporaneous crisis dummy yields similar results to standard specification; empirical support for within-year systematic spending response is weak.
- Omitting the OECD CLI from the first step:
  - Main results not sensitive, except for GDP and components under a peg.
  - As with stricter high-debt definition, fiscal multipliers appear smaller under a peg when CLI omitted.
  - Trade balance and exchange rate responses remain similar.
- Alternative sample time dimensions:
  - Excluding global financial crisis (limit sample to end-2006) and extending to early 1970s produce no appreciable differences.
- Excluding the United States:
  - Results broadly unchanged.
- Overall robustness conclusions:
  - Main findings robust across a wide range of definitions and specifications.
  - Real exchange rate response varies systematically with exchange rate regime.
  - Fiscal multipliers increase markedly during financial crises.
  - Policy implication: large conditional multipliers during financial crises support fiscal stimulus in acute financial stress; historical tendency to cut spending during crises implies large costs, arguing for building fiscal buffers in normal times.

*Source: _wp12150 - sections 2.1, 3.2, 5.1, 5.3 from the provided IMF content unit.*

### 2.1    A theoretical benchmark

### 2.1    A theoretical benchmark

### Private consumption response and implications for multipliers
- Modern business cycle models (neoclassical and new Keynesian) model private consumption as governed by intertemporal optimization and generally imply that private consumption falls in response to an increase in government spending.
- As a result, output multipliers are considerably smaller than predicted by traditional Keynesian analysis, which predicts a positive consumption response.
- Baxter and King (1993) find that impact multipliers on output hardly ever exceed unity across various specifications for household preferences and the duration of fiscal stimulus.
- Extensions and refinements have largely remained within standard business cycle models and often abstract from the exchange rate regime, government debt, and financial frictions.

### Pegged exchange rates
- In open economies with high capital mobility, the exchange rate regime determines the scope for independent monetary policy, central to fiscal policy analysis in the Mundell-Fleming framework.
- Textbook prediction: under flexible exchange rates, fiscal expansion crowds out net exports one-for-one via exchange rate appreciation (assuming unchanged monetary stance); under fixed exchange rates, fiscal policy is effective because monetary expansion offsets appreciation pressure.
- New Keynesian models temper these sharp predictions: fiscal stimulus can be either more or less effective under a flexible exchange rate depending on assumptions about monetary policy and the medium-term debt consolidation framework (Corsetti, Kuester, and M ̈uller (2011a)).
- Most standard models predict that higher government spending appreciates the real exchange rate.
- Exceptions and recent contributions identify conditions for the opposite:
  - Kollmann (2010): government spending may depreciate the real exchange rate if spending shocks are very persistent and international financial markets are incomplete.
  - Ravn, Schmitt-Groh ́e, and Uribe (2012): with “deep habits” in preferences, imperfect competition leads producers to lower markups and prices in the short run to lock in higher future public demand, causing the real exchange rate to depreciate.
- Corsetti, Meier, and M ̈uller (2011b): the impact of short-run fiscal stimulus (including on the exchange rate) crucially depends on the medium-term debt consolidation regime. If higher spending today contributes to debt stabilization via anticipated future spending cuts, long-term real interest rates need not rise and the real exchange rate can depreciate. This exchange rate response is driven by anticipated monetary policy responses to future spending cuts.

### Weak public finances
- Empirical observations (Giavazzi and Pagano (1990)) showed fiscal consolidations in Denmark and Ireland during the 1980s coincided with improving government balances and strong private consumption growth—puzzling from a Keynesian perspective but consistent with some neoclassical accounts.
- Literature identifies aspects of public finances that alter private consumption responses:
  - Bertola and Drazen (1993): correlation between private and public consumption changes as debt approaches a critical level that triggers sharp fiscal retrenchment.
  - Perotti (1999): model with unconstrained and credit-constrained households and distortionary taxes. Aggregate demand response to fiscal measures depends on the initial level of debt:
    - In “good” times, fiscal balances and private consumption comove negatively.
    - In “bad” times (high public debt), the comovement can become positive.
    - Mechanism: with high initial debt, tax distortions amplify negative wealth effects for unconstrained households, outweighing positive effects for constrained households.
- Perotti (1999) empirical approach (two-step):
  1. Estimate a near-VAR on annual data for 19 OECD countries to identify fiscal policy shocks.
  2. Regress consumption growth on the identified fiscal shocks and an interaction term with a “bad times” dummy (constructed from debt and deficit levels). Result: consumption responds positively to spending increases and tax cuts in normal times; sign reverses in bad times.
- Related findings: Sutherland (1997) finds tax cuts may be contractionary if debt is high and consolidation is expected during current generations’ lifetimes. Afonso and Jalles (2011) find reductions in government consumption tend to raise private consumption when public debt is high.

### Financial crises
- Financial crises restrict access to credit sharply; models that include a fraction of constrained households (who consume their disposable income each period) imply larger fiscal multipliers when the share of constrained households is higher (Gal ́ı et al. (2007); Bilbiie, Meier, and M ̈uller (2008); Fern ́andez-Villaverde (2010)).
- Corsetti et al. (2011a): credit constraints have a larger effect on multipliers under fixed exchange rates or in a monetary union.
- Financial crises can push monetary policy to the zero lower bound, impairing central bank stimulus and allowing fiscal stimulus to be fully accommodated at unchanged (zero) policy rates. This can halt deflationary dynamics caused by falling prices and rising real rates (Eggertsson and Woodford (2003)).
- Under zero lower bound conditions, fiscal multipliers on output can exceed two or three (Christiano et al. (2011); Hall (2009); Erceg and Lind ́e (2010); Woodford (2011)).
- Caveat: sovereign risk premia associated with higher public deficits can raise funding costs and may alter these multiplier results (Corsetti et al. (2012)).

### Empirical strategy (overview) and identification issues
- The paper pursues a two-stage estimation strategy similar to Perotti (1999):
  1. Estimate a fiscal policy rule for each country to describe the statistical process of government spending and obtain spending shocks. The rule links government spending to important macroeconomic aggregates and resembles the structure in fiscal policy VARs.
  2. Use the estimated policy shocks as regressors to trace impacts of government spending on output, private consumption, investment, the trade balance, and the real effective exchange rate.
- Motivation for this approach:
  - Standard structural VAR approaches (Blanchard and Perotti (2002) and related work) identify spending shocks by assuming discretionary government spending is subject to decision/implementation lags and thus unresponsive to contemporaneous conditions. This is plausible for government consumption and investment, which normally contain no automatic cyclical component.
  - Alternative approaches and their limitations:
    - Ramey and Shapiro (1998) event-based identification relies on a small number of arguably exogenous events; subsequent VAR implementations and forecast-error approaches (Ramey (2011)) highlight timing/anticipation issues.
    - Mountford and Uhlig (2009) use sign restrictions; Enders, M ̈uller, and Scholl (2011) derive sign restrictions from open-economy models.
    - Panel VARs (Ilzetzki et al. (2011)) allow subgroup analysis but are limited in isolating multiple dimensions simultaneously and in accounting for time-varying attributes (e.g., financial crises).
  - The chosen two-step approach permits:
    - Estimating fiscal policy rules for one country at a time to allow heterogeneity across national policymaking.
    - Flexible specification to account for differences in transmission across economic environments (pegged vs. flexible exchange rates; sound vs. strained fiscal positions; normal times vs. financial crisis).
- The fiscal policy rule-based first step is thus used to generate spending shocks that feed into the second-step regression analysis of macroeconomic responses.

*Source: _wp12150 - 2.1    A theoretical benchmark*

### 3.2    The first step: Identifying government spending shocks

### 3.2 The first step: Identifying government spending shocks

### Fiscal policy variable and rationale
- Policy variable: per capita government consumption expressed in logs.
- Government consumption accounts for a significant 21.5 percent of GDP in the average country in the sample.
- Rationale for using government consumption:
  - Contains virtually no automatic cyclical component.
  - No obvious direct link to private sector productivity (unlike public investment), limiting transmission channels.

### Empirical fiscal rule (specification)
- The process of government spending is modeled for each country i by:
  g_t,i = φ_i + η_i trend_t + β_i,1 g_{t−1,i} + β_i,2 g_{t−2,i} + γ_i,1 y_{t−1,i} + γ_i,2 y_{t−2,i} + θ_i cli_{t−1,i} + δ_i b_{t−1,i} + ρ_i,1 peg_{t−1,i} + ρ_i,2 strain_{t,i} + ρ_i,3 crisis_{t−1,i} + ε_{t,i}. (1)
- Features of the rule:
  - Includes first and second lags of government spending (g_{t−1}, g_{t−2}).
  - Includes first two lags of log per capita output (y_{t−1}, y_{t−2}).
  - Includes lagged composite leading indicator cli_{t−1} (normalized OECD CLI from October of the previous year) as a proxy for pre-budget growth expectations.
  - Includes beginning-of-period debt stock b_{t−1} (share of GDP).
  - Includes trend and constant.
  - Includes lagged-information dummies: exchange rate peg (peg_{t−1}), fiscal strain (strain_{t}), and financial crisis (crisis_{t−1}).
  - Parameters (φ_i, η_i, β_i, γ_i, θ_i, δ_i, ρ_i) are stable over time for each country but heterogeneous across countries.
  - Additive shock term ε_{t,i} captures unexpected discretionary policy changes (fiscal policy innovations).

### Identification assumptions and tests
- Key identifying assumption: no two-way contemporaneous interdependence between spending and its determinants (spending cannot respond to simultaneous output developments).
  - Spending responds to past growth developments and to expectations formed one period in advance.
  - Use of lagged CLI (October of the previous year) to capture expectations at budget formulation.
- Automatic stabilizers are not expected to bias identification because they operate mainly through (tax) revenue and transfers, not government consumption.
- Potential threats:
  - Discretionary policy responding to contemporaneous output developments — issue of precise definition of “contemporaneous.” Blanchard and Perotti (2002) argue limited within-quarter responsiveness; Born and Müller (2012) provide evidence supporting the restriction that spending does not respond within a year for several countries.
  - Anticipation of fiscal measures (Mountford and Uhlig (2009); Ramey (2011)) — this concern is greater for high-frequency (quarterly) data; annual data help attenuate foreseeability problems.
- Exceptional rapid fiscal responses (e.g., U.S. 2008–09 stimulus) are acknowledged; the specification includes a lagged financial crisis dummy to capture systematic fiscal responses during 2008–09. Sensitivity checks also limit sample to 2006 and earlier.
- Statistical adequacy:
  - Arellano-Bond tests for autoregression (lags 1, 2, 3) on country-specific residuals: the null hypothesis of no autocorrelation cannot be rejected at conventional levels for any country, supporting the residuals as measures of unanticipated spending changes.

### Definition of key dummies and related thresholds
- Exchange rate regime classification based on Ilzetzki et al. (2009).
- Financial crisis dates based on Reinhart and Rogoff (2008) and Reinhart (2010).
- Fiscal strain (weak public finances) defined as: government debt in excess of 100 percent of GDP or net government borrowing above 6 percent of GDP (each lagged once). These definitions are varied in robustness checks.

### Data and sample constraints relevant to identification
- Annual data used; maximum coverage 1975 through 2008 (robustness checks exclude the global financial crisis).
- Requirement: at least 20 consecutive annual observations to obtain reliable estimates of the fiscal policy rule in the first step.
- Final sample: 17 countries (due to data limitations): Australia, Austria, Belgium, Canada, Denmark, Finland, France, Ireland, Italy, Japan, Netherlands, Norway, Portugal, Spain, Sweden, the United Kingdom, and the United States.
- Final sample size for second-stage estimation: 444 country-year observations (reduced because of data gaps and deletions when, e.g., a country’s exchange rate regime changed within a year).

### Properties of the estimated shocks (first-step residuals)
- Standard deviation of retained shocks: 1.02 percent of government spending.
- Minimum and maximum shock values: -3.57 to 5.16 percentage points.
- Mean and median of shocks: close to zero.
- Correlation of estimated shocks with raw growth rate of government spending: 0.64 (indicating removal of systematic components while retaining resemblance to raw data).
- Interpretation caveat: interpreting shocks subject to caveats noted by Sims (1998).

### Qualitative results from country-specific fiscal-rule estimates
- Good fit of the empirical fiscal rule (high fit due to autoregressive terms in log levels); fit remains quite good when re-estimated in growth rates for robustness.
- General finding: government spending tends to respond negatively to the outstanding stock of public debt.
  - The coefficient on debt is negative for all but three countries (Austria, Finland, and the United Kingdom).
  - Negative coefficient is statistically significant for about half of the countries.
  - Implication: government spending plays a role in debt consolidation strategies; relates to Corsetti et al. (2011b) evidence on debt-responsive spending and spending reversals.
- Coefficients on lagged output and the composite leading indicator are harder to translate into a single statement about cyclical properties because related coefficients jointly determine cyclicality.
- Financial crisis dummy estimates:
  - For countries experiencing financial crises within the sample, the estimated crisis dummy coefficient is negative (and sometimes significantly so) in nine out of thirteen countries.
  - Interpretation: holding other factors fixed, government spending tended to slow down rather than accelerate during those financial crises — possibly reflecting financing constraints, banking sector bailouts, and falling tax revenues.

### Role of first-step shocks in the broader estimation strategy
- Estimated shocks (ε̂_{t,i}) serve as inputs for the second-step panel regressions that trace dynamic effects of government spending shocks on output, its components, international prices, and other macro variables.
- The specification and properties of the first-step shocks are tested and validated (no serial correlation; plausible distributional properties), providing a credible basis for the second-step causal analysis.

*Source: _wp12150 - 3.2 The first step: Identifying government spending shocks*

### 5.1    Unconditional effects

### 5.1    Unconditional effects

### Unconditional estimation approach
- Estimates obtained omitting all economic environment dummies listed in Table 5.
- Two-step procedure: first- and second-step regressions run without environment controls to produce unconditional impulse responses.
- Standard errors computed by drawing 1,000 realizations of the coefficient vector assuming a multivariate normal distribution with variance-covariance matrix corresponding to that of the regression coefficients; the mean is set equal to the point estimate.
- Impulse responses normalized so that the initial increase in government spending is one percent of GDP (notation: the point estimate of the response of government spending to the standardized shock tends to be around 0.9, reflecting a moderate attenuating bias in the second step caused by measurement error in the first step).

### Key unconditional findings (impulse responses; Figure 1)
- Government spending:
  - Persistent increase (normalized to one percent of GDP initial increase).
- Output:
  - Sizeable increase by about 0.7 percentage points on impact.
- Consumption:
  - Does not respond on impact (no immediate change).
- Investment:
  - Private investment falls (clear-cut negative response).
- Net exports:
  - Decline in net exports following the positive spending shock (supports notion of “twin deficits”).
- Inflation:
  - Drops on impact, then rises temporarily above baseline.
- Short-term interest rate:
  - Rises on impact and then moves in tandem with inflation.
- Real effective exchange rate:
  - Appreciates slightly on impact, but then depreciates over time.

### Notes on multipliers and comparison with literature
- Responses of output and consumption are measured in the same units as government spending; therefore responses represent estimates of government spending multipliers on output and consumption.
- Average multipliers found are within ranges commonly reported:
  - Output multipliers in the literature: range from 0.5 to 1.0 (Hall (2009) summary).
  - Consumption multipliers in the literature: range from somewhat negative to 0.5.
- Specific methodological notes from the source:
  - Blanchard-Perotti identification typically reports positive consumption multipliers (with some evidence of decline over time).
  - Ramey-Shapiro approach often reports negative consumption multipliers, but quantitatively small.
  - Mountford and Uhlig (2009), Barro and Redlick (2011), and Hall (2009) find no significant consumption response to deficit-financed or defense spending shocks.
- Alternative multiplier computation (cumulative response of variable divided by cumulative response of government spending) yields very similar results for the role of the economic environment.

### Interpretation and implications
- Unconditional results serve as a benchmark and align with existing empirical literature on fiscal multipliers.
- Negative investment and worsening trade balance are consistent with conventional wisdom and earlier empirical findings.
- The short-term decline in inflation after a government spending increase is surprising relative to standard theoretical predictions but has been documented under alternative identification schemes.
- Mixed findings on real exchange rate responses across studies may reflect differences in exchange rate regimes; unconditional results show slight initial appreciation followed by depreciation.
- The premise motivating the paper: unconditional models may conceal large differences in the response to fiscal shocks across economic environments — motivating the move to environment-conditioned analysis in subsequent sections.

_Italic: Source: _wp12150 - 5.1    Unconditional effects (PDF chapter/section) from the provided IMF content unit._

### 5.3    Sensitivity analysis

### 5.3    Sensitivity analysis

### Alternative definitions of financial crisis
- Narrower crisis dummy: includes only the “big five” banking crises (Finland 1991–94, Japan 1992–97, Norway 1988–93, Spain 1978–85, and Sweden 1991–94) plus the global financial crisis starting in 2007.
- Trade-off: selects crises with comparably large magnitude but reduces the number of observations considerably.
- Main robustness outcomes (relative to wider crisis definition):
  - Government spending appears less persistent.
  - Impact response for output and consumption remains close to 2.
  - Overall activity reverts to trend more quickly.
  - Responses of inflation and the real exchange rate are stronger, matched by a sharper rise in the short-term interest rate.
- Multipliers conditional on high debt are not affected by the narrower crisis definition.
- Under the peg, differences in impact responses of output and investment across currency regimes become larger.
- Note: the puzzling result of a delayed rise in consumption disappears in this specification.

### Alternative definition of weak public finances
- Stricter thresholds: debt raised to 120 percent of GDP and the deficit to 7 percent of GDP.
- Results:
  - Conditional on financial crisis and weak public finances: not substantially affected.
  - Under a peg, the point estimates of impact responses for output and investment are now lower relative to a regime of flexible exchange rates (reversal of prior main result).
  - External variables (real exchange rate and trade balance) show similar or stronger responses: the real exchange rate appreciates on impact, and the trade deficit worsens.

### Specification in differences (vs. levels)
- Model specified in differences; consider cumulative effect of a government spending shock.
- Robustness:
  - Results from the level specification are quite robust to the differences specification.
  - Unexpected spending increases remain considerably more effective during financial crises than in normal times.
  - Output and consumption multipliers remain around 2.
  - Investment rises over time.
  - Net exports deteriorate, and the real exchange rate depreciates.
  - Results conditional on weak public finances and an exchange rate peg are confirmed; under a peg, the output multiplier is more clearly in line with received wisdom.

### Contemporaneous crisis dummy (vs. lagged)
- Specification includes contemporaneous rather than lagged value of the crisis dummy in the first step, implying systematic within-year government spending response during crises.
- Empirical support for this hypothesis is weak, but results are quite similar to the standard specification.

### Omitting the OECD Composite Leading Indicator (CLI)
- Excluding lagged values of the CLI from the first step (CLI intended to capture advance information about cyclical developments).
- Rationale: Omitting CLI could confound anticipated cyclical spending changes with policy innovations; including CLI could induce spurious correlation if policymakers wrongly use CLI to stabilize.
- Findings:
  - Main results are not sensitive to omitting the CLI, except for GDP and its components under a peg.
  - As with the more restrictive high-debt definition, fiscal multipliers appear smaller under a peg when CLI is omitted.
  - Responses of the trade balance and the exchange rate remain similar to prior evidence.

### Alternative sample time dimensions
- Excluding global financial crisis: limit sample to end in 2006 to exclude observations from the global financial crisis.
  - Rationale: unusually short response time of fiscal policy during the latest episode could jeopardize identification.
- Extending sample: include observations from the early 1970s (late stages of Bretton Woods) in a separate unreported check.
- Result: neither experiment generates appreciable differences relative to the baseline sample; observations for the first two years of the global financial crisis do not appear influential.
- Conclusion: transmission of fiscal policy does not appear specific to the post-Bretton Woods international monetary system.

### Excluding the United States
- Re-estimation excluding the United States to test for a U.S. outlier effect.
- Result: results are broadly unchanged relative to the baseline setup.

### Overall robustness conclusions
- The main findings are robust to a wide range of alternative definitions and specifications:
  - Real exchange rate response varies systematically with exchange rate regime.
  - Fiscal multipliers increase markedly during financial crises.
- Policy-relevant implications reiterated:
  - Large conditional multipliers during financial crises support fiscal stimulus in times of acute financial stress.
  - Historical tendency of many countries to cut government spending during financial crises—presumably for debt sustainability reasons—implies large costs when retrenchment occurs in high-multiplier states.
  - This provides an argument for building up fiscal buffers in normal times to avoid painful fiscal retrenchment when financial turmoil strikes.

*Source: 5.3 Sensitivity analysis (excerpt).*

### References

### _wp12150 - References

### Key bibliographic references
- Acconcia, A., G. Corsetti, and S. Simonelli. "Mafia and public spending: Evidence of the fiscal multiplier from a quasi-experiment." mimeo, 2011.
- Afonso, A. and J. Jalles. "Assessing fiscal episodes." Technical university of Lisbon Working Paper 15/2011, 2011.
- Auerbach, A. and Y. Gorodnichenko. "Measuring the output responses to fiscal policy." American Economic Journal: Economic Policy, forthcoming, 2010.
- Auerbach, A. and Y. Gorodnichenko. "Fiscal multipliers in recession and expansion." NBER Working paper 17447, 2011.
- Backus, D., P. Kehoe, and F. Kydland. "Dynamics of the trade balance and the terms of trade: the J-curve?" American Economic Review, 84(1):84–103, March 1994.
- Baldacci, E., S. Gupta, and C. Mulas-Granados. "How effective is fiscal policy response in systemic banking crises?" IMF Working Paper WP/09/160, 2001.
- Barro, R. and C. J. Redlick. "Macroeconomic effects from government purchases and taxes." Quarterly Journal of Economics, 126:51–102, 2011.
- Baxter, M. and R. G. King. "Fiscal policy in general equilibrium." American Economic Review, 83(3):315–334, June 1993.
- Beetsma, R., M. Giuliodori, and F. Klaasen. "Trade spill-overs of fiscal policy in the European Union: a panel analysis." Economic Policy, 48:640–687, 2006.
- Beetsma, R., M. Giuliodori, and F. Klaasen. "The effects of public spending shocks on trade balances in the European Union." Journal of the European Economic Association, 6(2-3):414–423, 2008.
- Beetsma, R., M. Giuliodori, and F. Klaassen. "Temporal aggregation and svar identification, with an application to fiscal policy." Economics Letters, 105:253–255, 2009.
- Bénetrix, A. S. and P. R. Lane. "Fiscal shocks and the real exchange rate," 2009. IIIS Discussion Paper 286.
- Bertola, G. and A. Drazen. "Trigger points and budget cuts: Explaining the effects of fiscal austerity." American Economic Review, 83(1):11–26, 1993.
- Bilbiie, F. O., A. Meier, and G. J. Müller. "What accounts for the changes in U.S. fiscal policy transmission?" Journal of Money, Credit, and Banking, 40(7):1439–1469, 2008.
- Blanchard, O. J. and R. Perotti. "An empirical characterization of the dynamic effects of changes in government spending and taxes on output." Quarterly Journal of Economics, 117(4):1329–1368, November 2002.
- Born, B. and G. J. Müller. "Government spending in quarterly and annual time series." Journal of Money, Credit and Banking, 44(2-3):507–517, 2012.
- Burnside, C., M. Eichenbaum, and J. D. Fisher. "Fiscal shocks and their consequences." Journal of Economic Theory, 115:89–117, 2004.
- Canova, F. and E. Pappa. "Price differentials in monetary unions: The role of fiscal shocks." Economic Journal, 117:713–737, March 2007.
- Canzoneri, M. B., R. E. Cumby, and B. Diba. "New view on the transatlantic transmission of fiscal policy and macroeconomic policy coordination." In M. Buti, editor, Monetary and Fiscal Policies in EMU, pages 283–311. Cambridge University Press, 2003.
- Christiano, L., M. Eichenbaum, and C. Evans. "The effects of monetary policy shocks: Evidence from the flow of funds." Review of Economics and Statistics, 78(1):16–34, 1996.
- Christiano, L., M. Eichenbaum, and S. Rebelo. "When is the government spending multiplier large?" Journal of Political Economy, 119(1):78–121, 2011.
- Cimadomo, J. "Real-time data and fiscal policy analysis: a survey of the literature." Working Papers 11-25, Federal Reserve Bank of Philadelphia, 2011.
- Cogan, J. F., T. Cwik, J. B. Taylor, and V. Wieland. "New Keynesian versus old Keynesian government spending multipliers." Journal of Economic Dynamics and Control, 34, 2010.
- Corsetti, G. and G. J. Müller. "Twin deficits: Squaring theory, evidence and common sense." Economic Policy, 48:598–638, 2006.
- Corsetti, G., K. Kuester, and G. J. Müller. "Pegs, floats and the transmission of fiscal policy." CEPR Discussion paper 8180, 2011a.
- Corsetti, G., A. Meier, and G. J. Müller. "Fiscal stimulus with spending reversals." Review of Economics and Statistics, forthcoming, 2011b.
- Corsetti, G., K. Kuester, A. Meier, and G. J. Müller. "Sovereign risk, fiscal policy, and macroeconomic stability." IMF Working paper 12/33, 2012.
- Dornbusch, R. "Exchange rate economics: Where do we stand." Brookings Papers on Economic Activity, 1:143–185, 1980.
- Edelberg, W., M. Eichenbaum, and J. D. M. Fischer. "Understanding the effects of a shock to government purchases." Review of Economic Dynamics, 2:166–206, 1999.
- Eggertsson, G. B. and M. Woodford. "The zero interest-rate bound and optimal monetary policy." Brookings Papers on Economic Activity, 1:139–211, 2003.
- Enders, Z., G. J. Müller, and A. Scholl. "How do fiscal and technology shocks affect real exchange rates? New evidence for the U.S." Journal of International Economics, 83:53–69, 2011.
- Erceg, C. and J. Lindé. "Is there a fiscal free lunch in a liquidity trap?" CEPR Discussion Paper No. 7624, 2010.
- Erceg, C., C. Gust, and D. López-Salido. "The transmission of domestic shocks in open economies." In J. Gali and M. Gerlter, editors, International Dimensions of Monetary Policy. University of Chicago Press, 2010.
- Fernández-Villaverde, J. "Fiscal policy in a model with financial frictions." American Economic Review: Papers & Proceedings, 100:35–40, May 2010.
- Galí, J. and R. Perotti. "Fiscal policy and monetary integration in Europe." Economic Policy, 37:534–572, 2003.
- Galí, J., J. D. López-Salido, and J. Vallés. "Understanding the effects of government spending on consumption." Journal of the European Economic Association, pages 227–270, 2007.
- Giavazzi, F. and M. Pagano. "Can severe fiscal contractions be expansionary? Tales of two small European countries." NBER Macroeconomics Annual, 5:75–111, 1990.
- Giavazzi, F., T. Jappelli, and M. Pagano. "Searching for non-linear effects of fiscal policy: Evidence from industrial and developing countries." European Economic Review, 44:1259–1289, 2000.
- Hall, R. E. "By How Much Does GDP Rise if the Government Buys More Output?" Brookings Papers on Economic Activity, (2):183–231, 2009.
- Hamilton, J. D. Time Series Analysis. Princeton University Press, 1994.
- Ilzetzki, E. O., C. Reinhart, and K. Rogoff. "Exchange rate arrangements entering the 21st century: Which anchor will hold?" 2009. Mimeo.
- Ilzetzki, E. O., E. Mendoza, and C. Vegh. "How big (small?) are fiscal multipliers?" 2011. mimeo.
- International Monetary Fund. "From recession to recovery: How soon and how strong?" World Economic Outlook, April:103–138, 2009.
- Kim, S. and N. Roubini. "Twin deficit or twin divergence? Fiscal policy, current account, and real exchange rate in the U.S." Journal of International Economics, 74(2):362–383, 2008.
- Kollmann, R. "Government purchases and the real exchange rate." Open Economies Review, 21:49–64, 2010.
- Lane, P. R. and R. Perotti. "The importance of composition of fiscal policy: Evidence from different exchange rate regimes." Journal of Public Economics, 87:2253–2279, 2003.
- Linnemann, L. and A. Schabert. "Fiscal policy and the new neoclassical synthesis." Journal of Money, Credit and Banking, 35(6):911–929, 2003.
- Monacelli, T. and R. Perotti. "Fiscal policy, the real exchange rate and traded goods." Economic Journal, 120:437–461, 2010.
- Mountford, A. and H. Uhlig. "What are the effects of fiscal policy shocks?" Journal of Applied Econometrics, 24(6):960–992, 2009.
- Newey, W. and D. McFadden. "Large sample estimation and hypothesis testing." In R. Engle and D. McFadden, editors, Handbook of Econometrics, volume 4, chapter 36, pages 2111–2245. Elsevier, 1994.
- Obstfeld, M. and K. Rogoff. "Exchange rate dynamics redux." Journal of Political Economy, 103:624–660, 1995.
- Perotti, R. "Fiscal policy in good times and bad." Quarterly Journal of Economics, 114(4):1399–1436, November 1999.
- Perotti, R. "Estimating the effects of fiscal policy in OECD countries." mimeo, 2004.
- Perotti, R. "In search of the transmission mechanism of fiscal policy." NBER Working paper 13143, 2007.
- Ramey, V. A. "Identifying government spending shocks: It‘s all in the timing." Quarterly Journal of Economics, 126(1):1–50, 2011.
- Ramey, V. A. and M. D. Shapiro. "Costly capital reallocation and the effects of government spending." Carnegie Rochester Conference on Public Policy, 48:154–194, 1998.
- Ravn, M. O., S. Schmitt-Grohè, and M. Uribe. "Consumption, government spending, and the real exchange rate." Journal of Monetary Economics, 59:215–234, 2012.
- Reinhart, C. "This time is different chartbook: Country histories on debt, default, and financial crisis." NBER Working Paper 15815, 2010.
- Reinhart, C. and K. Rogoff. "Banking crises: An equal opportunity menace," 2008. NBER Working Paper 14587.
- Ricci, L. A., G. M. Milesi-Ferretti, and J. Lee. "Real exchange rates and fundamentals: A cross-country perspective." IMF Working Paper 08/13, 2008.
- Rotemberg, J. and M. Woodford. "Oligopolistic pricing and the effects of aggregate demand on economic activity." Journal of Political Economy, 100(6):1153–1207, 1992.
- Sims, C. "Comment on Glenn Rudebusch’s 'Do measures of monetary policy in a var make sense?'." International Economic Review, 39:933–941, 1998.
- Sutherland, A. "Fiscal crisis and aggregate demand: Can high public debt reverse the effects of fiscal policy?" Journal of Public Economics, 65:147–162, 1997.
- Tagkalakis, A. "The effects of fiscal policy on consumption in recessions and expansions." Journal of Public Economics, 92:1486–1508, 2008.
- Woodford, M. "Simple analytics of the government expenditure multiplier." American Economic Journal: Macroeconomics, 3:1–35, 2011.

### Tables, definitions, and data sources
- Table 1. Composition of Initial and Final Samples
  - Initial sample time window: maximum time period 1975-2008, depending on data availability; initial sample comprises the same set of OECD countries considered in Perotti (1999).
  - Total no. of observations: 517444
  - 2/ Final sample determined by availability of all regressors included in the most comprehensive specification for second-step regression.
- Table 2. Data Sources and Definitions — selected variables and sources:
  - Government spending: Log of real per capita government consumption. OECD Economic Outlook Database: volume of final government consumption expenditure (CGV); OECD Analytic Database: population size (POP).
  - GDP: Log of per capita GDP. OECD Economic Outlook Database: value of gross domestic product (GDP), GDP deflator (PGDP); OECD Analytic Database: population size (POP).
  - CLI: Composite leading indicator. OECD Main Economic Indicators database: CLI amplitude-adjusted; normalized by subtracting 100, and dividing by 100.
  - Public debt: General government gross debt (in percent of GDP). Primary source: IMF World Economic Outlook: General government gross debt (GGD), nominal GDP (NGDP); where unavailable: OECD Analytic Database: General government gross financial liabilities as a percentage of GDP (GGFLQ); for Sweden: Statistics Sweden: Central government gross debt in percent of GDP.
  - Financial crisis dummy: Takes on value of 1 during financial crises, and 0 otherwise. Reinhart and Rogoff (2008, pp. 65 ff.) and Reinhart and Rogoff (2010): the relevant crisis episodes, broadly defined, are (severe financial crises in bold): Australia 1989-92; Austria 2008; Belgium 2008; Canada 1983-85; Denmark 1987-92 and 2008; Finland 1991-94; France 1994-95; Germany 2007-08; Ireland 2007-08; Italy 1990-94 and 2008; Japan 1992-97; Netherlands 2008; Norway 1988-93; Spain 1978-85 and 2008; Sweden 1991-94; United Kingdom 1974-76, 1984, and 2007-08; United States 1984-91 and 2007-08.
  - Bad fiscal times dummy: Takes on value of 1 when lagged public debt exceeds 100 percent of GDP or lagged government net borrowing exceeds 6 percent of GDP, and 0 otherwise. OECD Analytic Database: Government net lending as a percentage of GDP (NLGQ); see also entry for public debt above.
  - Peg dummy: Takes on value of 1 when exchange rate regime defined as peg, and 0 otherwise. Ilzetzki, Reinhart, and Rogoff (2008): categories 1-8 of the authors' fine classification scheme are defined as "peg"; category 14 ("freely falling") coded as "not available"; within-year changes in regime category also lead to coding as "not available"; time series updated for 2008 (identical classification as for 2007) by the authors.
  - Private consumption: Log per capita real private consumption. OECD Economic Outlook Database: volume of final private consumption expenditure (CPV); OECD Analytic Database: population size (POP).
  - Private investment: Log per capita real fixed investment. OECD Economic Outlook Database: volume of private total fixed capital formation (IPV); OECD Analytic Database: population size (POP).
  - Trade balance: Ratio of net exports to GDP. IMF World Economic Outlook: exports of goods and services at current prices (NX), imports of goods and services at current prices (NM), nominal GDP (NGDP).
  - REER: CPI-based real effective exchange rate (in percent). OECD Main Economic Indicators Database (CCRETT01.IXOB).
  - Interest rate: Nominal short-term interest rate. IMF International Financial Statistics: money market rate.
  - Inflation: Annual change in Consumer Price Index (all items), average for the year. OECD Main Economic Indicators Database (CPALTT01.GY).

### Sample composition (selected entries)
- Country initial/final sample periods (examples):
  - Australia: 1989-2008 / 1992-2008
  - Austria: 1975-2008 / 1978-2008
  - Belgium: 1975-2008 / 1978-2008
  - Canada: 1975-2002, 2004-08 / 1978-2001
  - USA: 1975-78, 1980-2008 / 1983-2008

### Summary statistics for estimated government spending shocks (final sample) (Percent)
- No. of observations: 444
- Mean: 0.04
- Median: 0.00
- Standard deviation: 1.02
- Minimum: -3.57
- Maximum: 5.16
- Correlation with simple growth rate of government spending: 0.64
- Five largest negative shocks:
  - Portugal, 1993: -3.57
  - Netherlands, 1984: -3.33
  - Netherland, 2005: -3.18
  - Norway, 1988: -2.97
  - Spain, 1988: -2.67
- Five largest positive shocks:
  - Portugal, 1991: 2.60
  - Portugal, 2005: 2.68
  - Denmark, 1993: 2.85
  - Ireland, 1986: 3.83
  - Netherlands, 2006: 5.16
- Note: 1/ All statistics refer to final sample described in Table 1.

### Dummy characteristics (Table 5) — countries and time periods fulfilling criteria in the final sample
- Peg (examples):
  - Austria, 1978-2008
  - Belgium, 1978-2008
  - Canada, 1978-2001
  - Denmark, 1978-2008
  - France, 1982-2008
  - Netherlands, 1978-2008
  - Spain, 1984-2008
  - Sweden, 1978-92
- Bad fiscal times (examples):
  - Belgium, 1978-2003
  - Canada, 1983-87, 1992-97
  - Denmark, 1982-84
  - France, 1994
  - Ireland, 1983-89
  - Italy, 1980-91, 1998-2008
  - Japan, 1997-2008
- Financial crisis (examples):
  - Australia, 1992
  - Austria, 2008
  - Belgium, 2008
  - Canada, 1983-85
  - Denmark, 1987-92, 2008
  - Finland, 1991
  - Ireland, 2007-08
  - Italy, 1990-91, 2008
  - Japan, 1992-97
  - Netherlands, 2008
  - Norway, 1988-93
  - Spain, 1984-85, 2008
  - Sweden, 1991-94
  - United Kingdom, 2007-08
  - United States, 1984-91, 2007-08

### Figures and robustness tests (captions and notes)
- Figure 5: Results for narrow definition of financial crisis. Notes: quantities measured in output units; solid line: point estimate for baseline scenario (shaded area: ± one standard error); dashed dotted line (dashed lines: ± one standard error): deviation from baseline under the specific condition indicated on top of column.
- Figure 6: Results for alternative definition of weak public finances (government debt >120 percent of GDP and/or lagged net borrowing >7 percent of GDP). Notes: see Figure 5.
- Figure 7: Results for difference specification. Notes: responses display cumulative effect of government spending innovation for model estimated in first differences; quantities measured in output units; solid line: point estimate for baseline scenario (shaded area: ± one standard error); dashed dotted line (dashed lines: ± one standard error): deviation from baseline under specific condition indicated on top of column.
- Figure 8: Results for first-step specification which includes contemporaneous value of crisis dummy. Notes: see Figure 5.
- Figure 9: Results without CLI in first step. Notes: see Figure 5.
- Figure 10: Results for sample without 2007–2008. Notes: quantities measured in output units; solid line: point estimate for baseline scenario (shaded area: ± one standard error); dashed dotted line (dashed lines: ± one standard error): deviation from baseline under specific condition indicated on top of column.
- Figure 11: Results for sample without United States. Notes: quantities measured in output units; solid line: point estimate for baseline scenario (shaded area: ± one standard error); dashed dotted line (dashed lines: ± one standard error): deviation from baseline under specific condition indicated on top of column.

*Source: _wp12150 - References*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12150.pdf_