## Output Losses in Europe during COVID-19: What Role for Policies?

## Source details

**Canonical URL:** [Output Losses in Europe during COVID-19: What Role for Policies?](https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022130-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2022/english/wpiea2022130-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2022/english/wpiea2022130-print-pdf.pdf.json)

---

### I. Scope and research questions
- Sample: 43 European countries (including 27 advanced and 16 emerging economies).
- Main questions:
  - What explains heterogeneity in 2020 growth outcomes?
  - Role of sectoral composition, country fundamentals, and macroeconomic policies?
  - Efficacy of different policy instruments?

### II. Key data and stylized facts
- Data sources and indicators:
  - Quarterly data from 2020; main macro series from IMF World Economic Outlook.
  - Sectoral GVAs from Eurostat or national sources; shadow economy share from Medina and Schneider (2018).
  - Mobility: Oxford stringency index and a constructed de facto mobility indicator from Google mobility residuals.
- Stylized contrasts and key statistics:
  - Average 2020 real GDP contractions: emerging economies in Europe contracted by 2.1 percent, advanced economies contracted by 6.7 percent.
  - Sectoral impacts: wholesale and retail were the largest contributors to the recession in nearly all countries; ICT helped mitigate contractions.
  - Policy responses: fiscal support substantially larger in advanced countries; emerging economies cut policy rates further while advanced economies relied more on unconventional monetary instruments.

### III. Decomposition exercise — Methodology
- Decomposition layers for 2020 annual growth:
  1. Underlying growth momentum: g1 ≡ g2020 − g* (g* proxied by October 2019 WEO growth projections).
  2. Sectoral composition (N = 10 sectors): quantify contribution g2 by benchmarking each country against the PPP-weighted average sectoral weights of European countries (w̃i).
  3. Within-sector losses: panel regressions by sector i over 2020Q1–Q4:
     - (g2020,i,c,t − g i,c,t *) = αi + βi Xc + γi Pc,t−1 + φi Mc,t + εi,c,t
     - Xc: vector of pre-pandemic fundamentals (standardized).
     - Pc,t−1: lagged policy variables (not standardized).
     - Mc,t: contemporaneous mobility variables (not standardized).
- Aggregation and unexplained component:
  - Contributions aggregated using counterfactual sector weights w̃i to make sectoral-composition and within-sector effects separable.
  - Unexplained portion: ∑ w̃i (αi + εi,c,t).

### IV. Decomposition exercise — Main results
- Single largest contributor:
  - The decline in mobility is the single largest contributor to output losses in all countries.
- Underlying growth momentum:
  - Positive contribution of underlying growth in many emerging market economies helped limit observed GDP contractions.
- Sectoral composition:
  - Negative contribution for tourism-dependent economies (e.g., Croatia, Spain, Greece) but overall limited quantitative role for most countries given PPP-weighted benchmarking and 10-sector aggregation.
- Initial conditions:
  - Demographic and health factors (e.g., lower median age) contributed significantly to greater resilience in emerging markets.
  - Higher informality also contributed positively to growth differentials.
  - Higher pre-pandemic current account surpluses in advanced economies somewhat offset emerging economy advantages.
- Policy contributions and caveats:
  - Policies (fiscal and monetary) helped mitigate output losses in all countries, with larger measured contributions in advanced economies reflecting larger policy packages.
  - Regression-based policy contributions likely biased downward (reverse causality, omitted variables, anticipation effects, inability to capture synchronized global easing); thus estimated contributions are lower bounds.

### V. Calibration exercise — Methodology
- Purpose: address downward bias in regression estimates via calibration.
- Fiscal multiplier specification:
  - Fc,t = MATL PATL,t + MLIQ PLIQ,t + MBTL PBTL,t
  - PATL, PLIQ, PBTL are shares of above-the-line, liquidity, and below-the-line measures (PBTL = 1 − PATL − PLIQ).
- Calibrated multiplier parameter values:
  - MATL = 0.83
  - MLIQ = 0.45
  - MBTL = τ MLIQ with τ = 1/3
- Monetary calibration:
  - Separates policy rate cuts and central bank balance sheet expansion; multipliers differentiated across countries where literature permits.
- Aggregate annualized policy contribution:
  - Sum over four quarters of fiscal contribution (fiscal measures × Fc,t) plus policy rate and balance sheet contributions.

### VI. Calibration exercise — Results
- Calibration impact on estimated policy contributions:
  - Calibration raises the potential contribution of policies by over 70 percent in advanced economies.
  - Calibration more than doubles the potential policy contribution in emerging market economies.

### VII. Determinants of fiscal multipliers — Empirical approach
- Estimation:
  - Country-level panel regressions for 2020 GDP performance (net of underlying growth and sectoral composition).
  - Fiscal support measures normalized by pre-pandemic GDP; lagged to reduce reverse causality.
  - Country fixed effects included; regressions start in 2020Q2.
  - Tests for non-linearities via quadratic terms for fiscal measures.

### VIII. Determinants of fiscal multipliers — Key findings and statistics
- Average fiscal package magnitudes (percent of pre-crisis GDP):
  - Above-the-line: 0.7 percent (advanced) vs 0.4 percent (emerging).
  - Liquidity measures: 0.2 percent (advanced) vs 0.03 percent (emerging).
  - Other fiscal measures: 2.5 percent (advanced) vs 0.5 percent (emerging).
- Composition shares (sample averages):
  - Above-the-line: 56.9 percent of all fiscal measures.
  - Other fiscal measures: 34.4 percent.
  - Liquidity measures: 8.7 percent.
  - Emerging economies relied more on above-the-line measures: 69 percent vs 50 percent in advanced economies.
  - Emerging economies relied less on other measures: 27 percent vs 38.5 percent in advanced economies.
  - Emerging economies relied less on liquidity measures: 3.8 percent vs 11.8 percent in advanced economies.
- Non-linearities / diminishing returns:
  - Total fiscal package shows non-linear effect: linear coefficient 0.404*** and quadratic term −0.021***.
  - Interpretation: fiscal multipliers decline as the size of the fiscal package increases.
- Multipliers by type (selected regression coefficients and robust p-values):
  - Above-the-line (ATL):
    - Linear-only: 1.032*** (0.000)
    - With quadratic: 2.413*** (0.000) and quadratic term −0.202*** (0.001)
    - Implication: ATL measures have the highest multiplier (above one).
  - Liquidity measures:
    - Linear-only: 0.182 (0.570) — not statistically significant.
    - With quadratic: 0.899 (0.160) and quadratic term −0.212** (0.027)
  - Other fiscal measures:
    - Linear-only: 0.218** (0.042)
    - With quadratic: 0.493*** (0.000) and quadratic term −0.013*** (0.001)
- Regression summary statistics:
  - Observations: 123
  - R-squared ranges: 0.784 to 0.865 depending on specification
  - Country fixed effects: YES; Time fixed effects: NO
  - Significance notation: *** p<0.01, ** p<0.05, * p<0.1
- Interpretation:
  - Above-the-line measures (additional spending and forgone revenue, direct transfers, wage subsidies, health spending, unemployment benefits) deliver the largest GDP impact per unit of fiscal support.
  - Other fiscal measures (below-the-line: equity injections, asset purchases, loans, guarantees) have positive but much smaller multipliers.
  - Liquidity measures show ambiguous statistical significance; non-linear term suggests diminishing returns where they matter.
  - Diminishing returns imply that smaller fiscal packages can exhibit higher marginal multipliers, explaining in part larger estimated multipliers in emerging economies.

### IX. Inequality, informality, and IMF-supported programs — Findings and statistics
- Inequality and fiscal effectiveness:
  - Cross-country variation example: "31.3 for advanced economies. It also varies between countries, ranging from 24.8 in Slovenia to 42 in Turkey."
  - Theoretical ambiguity: higher inequality can increase or decrease fiscal multipliers depending on distribution of liquidity-constrained households.
  - Empirical finding (Panel A, Table 2): higher levels of inequality are associated with a higher fiscal multiplier when non-linearities are accounted for.
  - Selected regression coefficients (Panel A highlights):
    - Total fiscal package coefficients: "0.404***", "0.948***", "0.985***", "0.922***", "0.767***" with p-values "(0.000)".
    - Total fiscal package, quadratic term: "-0.021***", "-0.023***", "-0.028***", "-0.025***" with p-values "(0.000)".
    - Fiscal package * Inequality (quartiles): "0.325***" with p-value "(0.002)".
  - Observations: 123; R-squared examples: "0.784", "0.842", "0.849", "0.863", "0.853".
- Informality and fiscal efficacy:
  - Average informality (share of GDP, 2015–17):
    - Emerging countries: "29.3 percent of GDP"
    - Advanced economies: "14.8 percent"
  - Channels: informality can both decrease and increase fiscal multipliers depending on access to supports and enforcement of containment.
  - Empirical finding (Panel B, Table 2): interactive term between informality and fiscal policy support is positive, but significant only in some specifications; results suggest a weak positive association.
  - Selected coefficients (Panel B highlights): total fiscal package coefficients: "0.491***", "0.352***", "0.194", "0.964***", "0.897***", "0.790***".
  - Fiscal * Informality (level) examples: "0.180*" and "0.107*" (weaker significance).
  - Observations: 123; R-squared examples: "0.798", "0.789", "0.847", "0.846".
- Role of IMF-supported programs:
  - IMF assistance during 2020: "Albania, Moldova, Ukraine, Bosnia and Herzegovina and Kosovo received IMF funding in 2020 through IMF-supported programs."
  - Mechanisms: IMF financing may relax fiscal space constraints, catalyze capital inflows, and enhance policy design and targeting.
  - Empirical test (Table 3 interaction): interaction of fiscal policy support with a dummy for IMF-supported program in 2020 yields a significant positive coefficient, "indicating that the fiscal multiplier is higher in countries with IMF programs."
  - Caveats: selection bias and concurrent disbursements from other official creditors may confound interpretation.
  - Table A6 selected interaction coefficients (Fiscal * IMF): examples include 1.624** (col 1), 1.584** (col 2), 1.891** (col 3), 1.174*** (col 4), 2.817* (col 10) alongside some negative or insignificant entries in other columns.
  - Observations: 123; R-squared ranges: 0.784 to 0.876.

### X. Differences in marginal fiscal multipliers and decomposition
- Marginal multiplier computation:
  - Country-by-country marginal fiscal multiplier computed using regression coefficients and country-specific data on inequality, informality, IMF programs, fiscal policy magnitudes and compositions; effect set to 0 if quadratic term dominates linear term.
- Stylized results:
  - Figures show larger marginal multipliers for additional fiscal support in emerging countries compared to advanced countries.
  - Explanatory factors for larger marginal multipliers in emerging countries: higher share of above-the-line measures, smaller fiscal packages, higher inequality, larger shadow economy, and presence of IMF-supported programs.
  - Despite larger estimated multipliers in emerging countries, the role of policies in advanced economies was much more important in mitigating the crisis due to the considerably larger size of announced policies in advanced economies.
  - Calibration/decomposition reiteration: policy measures raised the potential contribution of policies by "over 70 percent in advanced economies and more than doubling it in emerging economies."

### XI. Overall findings and policy implications
- Main drivers of cross-country growth differentials during 2020:
  - Decline in mobility is the largest single contributor to output losses in all countries.
  - Differences in underlying growth trends, pre-pandemic fundamentals (demographic, health, informality), and macroeconomic policies explain much heterogeneity.
  - Sectoral composition played a limited role in explaining cross-country differences within Europe given PPP-weighted benchmarking and 10-sector aggregation.
- Policy effectiveness:
  - Fiscal and monetary policies helped cushion the pandemic shock; calibrated estimates imply substantially larger effects than regression estimates.
  - Above-the-line fiscal measures are most effective per unit of GDP, but multipliers exhibit non-linear diminishing returns as package size increases.
- Policy implications:
  - Composition matters: prioritize above-the-line measures (targeted transfers, wage support, health spending, unemployment benefits) to maximize GDP impact per unit of support.
  - Consider country characteristics: inequality, informality, and access to international financing (including IMF programs) affect marginal fiscal effectiveness.
  - Be mindful of diminishing returns: marginal effectiveness falls as fiscal packages become larger.
- Important caveats:
  - Potential endogeneity, omitted variable bias, and anticipation effects bias empirical estimates downward.
  - Cross-country methodology cannot fully capture synchronized global easing and international financial spillovers.
  - Analysis covers only 2020 and does not account for the role of vaccinations in the recovery.

*Italic: IMF Working Paper — "Output Losses in Europe during COVID-19: What Role for Policies?"*

### References .............................................................................................................

### Output Losses in Europe during COVID-19: What Role for Policies?

### I. Scope and research questions
- Sample: 43 European countries (including 27 advanced and 16 emerging economies).
- Main questions:
  - What explains heterogeneity in 2020 growth outcomes?
  - Role of sectoral composition, country fundamentals, and macroeconomic policies?
  - Efficacy of different policy instruments?

### II. Key data and stylized facts
- Quarterly data from 2020; main macro series from IMF World Economic Outlook.
- Sectoral GVAs from Eurostat or national sources; shadow economy share from Medina and Schneider (2018).
- Mobility: Oxford stringency index and a constructed de facto mobility indicator from Google mobility residuals.
- Stylized contrasts:
  - Average 2020 real GDP contractions: emerging economies in Europe contracted by 2.1 percent, advanced economies contracted by 6.7 percent.
  - Sectoral impacts: wholesale and retail were the largest contributors to the recession in nearly all countries; ICT helped mitigate contractions.
  - Policy responses: fiscal support substantially larger in advanced countries; emerging economies cut policy rates further while advanced economies relied more on unconventional monetary instruments.

### III. Decomposition exercise — Methodology
- Decomposition layers for 2020 annual growth:
  1. Underlying growth momentum: g1 ≡ g2020 − g* (g* proxied by October 2019 WEO growth projections).
  2. Sectoral composition (N = 10 sectors): quantify contribution g2 by benchmarking each country against the PPP-weighted average sectoral weights of European countries (w̃i).
  3. Within-sector losses: panel regressions by sector i over 2020Q1–Q4:
     - (g2020,i,c,t − g i,c,t *) = αi + βi Xc + γi Pc,t−1 + φi Mc,t + εi,c,t
     - Xc: vector of pre-pandemic fundamentals (standardized).
     - Pc,t−1: lagged policy variables (not standardized).
     - Mc,t: contemporaneous mobility variables (not standardized).
- Contributions aggregated using counterfactual sector weights w̃i to make sectoral-composition and within-sector effects separable.
- Unexplained portion: ∑ w̃i (αi + εi,c,t).

### IV. Decomposition exercise — Main results
- Single largest contributor to output losses in all countries: the decline in mobility.
- Underlying growth momentum:
  - Positive contribution of underlying growth in many emerging market economies helped limit observed GDP contractions.
- Sectoral composition:
  - Negative contribution for tourism-dependent economies (e.g., Croatia, Spain, Greece) but overall limited quantitative role for most countries.
  - Explanations for limited sectoral role include benchmarking against European averages, offsetting correlations across sectors, a 10-sector aggregation, and non-capture of cross-sector spillovers.
- Initial conditions:
  - Demographic and health factors (e.g., lower median age) contributed significantly to greater resilience in emerging markets.
  - Higher informality also contributed positively to growth differentials.
  - Higher pre-pandemic current account surpluses in advanced economies somewhat offset emerging economy advantages.
- Policy contributions:
  - Policies (fiscal and monetary) helped mitigate output losses in all countries, with larger measured contributions in advanced economies reflecting larger policy packages.
  - Regression-based policy contributions likely biased downward (reverse causality, omitted variables, anticipation effects, inability to capture synchronized global easing); thus estimated contributions are lower bounds.

### V. Calibration exercise — Methodology
- Address downward bias by calibrating fiscal multipliers based on composition of announced fiscal measures from the IMF COVID-19 Policy Survey.
- Fiscal multiplier for country c and quarter t:
  - Fc,t = MATL PATL,t + MLIQ PLIQ,t + MBTL PBTL,t
  - PATL, PLIQ, PBTL are shares of above-the-line, liquidity, and below-the-line measures (PBTL = 1 − PATL − PLIQ).
- Calibrated multipliers:
  - MATL = 0.83 (average from Bayer et al. 2020; Guerrieri et al. 2020; Faria-e-Castro 2021).
  - MLIQ = 0.45 (Faria-e-Castro 2021).
  - MBTL = τ MLIQ with τ = 1/3 (low take-up assumption for below-the-line measures).
- Monetary policy calibration separates policy rate cuts and central bank balance sheet expansion; multipliers differentiated across countries where literature permits.
- Annualized country policy contribution = sum over four quarters of fiscal contribution (fiscal measures × Fc,t) plus policy rate and balance sheet contributions.

### VI. Calibration exercise — Results
- Calibrated contributions of announced measures are substantially larger than regression estimates:
  - Calibration raises the potential contribution of policies by over 70 percent in advanced economies.
  - Calibration more than doubles the potential policy contribution in emerging market economies.

### VII. Determinants of fiscal multipliers — Empirical approach
- Country-level panel regressions for 2020 GDP performance (net of underlying growth and sectoral composition).
- Fiscal support measures (above-the-line, liquidity, other) normalized by pre-pandemic GDP; lagged to reduce reverse causality.
- Country fixed effects included; regressions start in 2020Q2.
- Tests for non-linearities by including quadratic terms for fiscal measures.

### VIII. Determinants of fiscal multipliers — Key findings
- Fiscal packages: average magnitudes (advanced vs emerging, percent of pre-crisis GDP):
  - Above-the-line: 0.7 percent (advanced) vs 0.4 percent (emerging).
  - Liquidity measures: 0.2 percent (advanced) vs 0.03 percent (emerging).
  - Other fiscal measures: 2.5 percent (advanced) vs 0.5 percent (emerging).
- Composition shares (sample averages):
  - Above-the-line: 56.9 percent of all fiscal measures.
  - Other fiscal measures: 34.4 percent.
  - Liquidity measures: 8.7 percent.
  - Emerging economies relied more on above-the-line measures: 69 percent vs 50 percent in advanced economies.
  - Emerging economies relied less on other measures: 27 percent vs 38.5 percent in advanced economies.
  - Emerging economies relied less on liquidity measures: 3.8 percent vs 11.8 percent in advanced economies.
- Non-linearities / diminishing returns:
  - Total fiscal package shows non-linear effect: linear coefficient 0.404*** and quadratic term −0.021*** (robust pvals in parentheses).
  - Interpretation: fiscal multipliers decline as the size of the fiscal package increases.
- Multipliers by type (Table 1 coefficients, robust pvals in parentheses):
  - Above-the-line (ATL) measures:
    - Linear-only: 1.032*** (0.000)
    - With quadratic: 2.413*** (0.000) and quadratic term −0.202*** (0.001)
    - Implication: ATL measures have the highest multiplier (above one).
  - Liquidity measures:
    - Linear-only: 0.182 (0.570) — not statistically significant.
    - With quadratic: 0.899 (0.160) and quadratic term −0.212** (0.027)
  - Other fiscal measures:
    - Linear-only: 0.218** (0.042)
    - With quadratic: 0.493*** (0.000) and quadratic term −0.013*** (0.001)
  - Regression summary statistics:
    - Observations: 123
    - R-squared ranges: 0.784 to 0.865 depending on specification
    - Country fixed effects: YES; Time fixed effects: NO
    - Significance notation: *** p<0.01, ** p<0.05, * p<0.1
- Interpretation:
  - Above-the-line measures (additional spending and forgone revenue, direct transfers, wage subsidies, health spending, unemployment benefits) deliver the largest GDP impact per unit of fiscal support.
  - Other fiscal measures (below-the-line: equity injections, asset purchases, loans, guarantees) have positive but much smaller multipliers.
  - Liquidity measures show ambiguous statistical significance; non-linear term suggests diminishing returns where they matter.
  - Diminishing returns imply that smaller fiscal packages can exhibit higher marginal multipliers, explaining in part larger estimated multipliers in emerging economies.
- Role of inequality and informality (introduced through interactions with fiscal measures; preliminary descriptive result):
  - The paper explores whether higher inequality and informality amplify fiscal effectiveness by interacting these measures with fiscal support. (Detailed interaction results continue beyond provided excerpt.)

### IX. Overall findings and caveats
- Overall explanatory factors for 2020 growth differentials:
  - Decline in mobility is the largest single contributor to output losses in all countries.
  - Differences in underlying growth trends, pre-pandemic fundamentals (demographic, health, informality), and macroeconomic policies explain much of the heterogeneity across countries.
  - Sectoral composition played a limited role in explaining cross-country differences within Europe given the PPP-weighted benchmarking approach and 10-sector aggregation.
- Policy impact:
  - Fiscal and monetary policies played an important role in cushioning the impact of the pandemic; calibrated estimates suggest substantially larger effects than regression estimates.
  - Above-the-line fiscal measures are most effective per unit of GDP, with non-linear diminishing returns.
- Important caveats:
  - Potential endogeneity, omitted variable bias, and anticipation effects bias empirical estimates downward.
  - Cross-country methodology cannot fully capture synchronized global easing and international financial spillovers.
  - Analysis covers only 2020 and thus does not account for the role of vaccinations in the recovery.

*Italic: IMF Working Paper — "Output Losses in Europe during COVID-19: What Role for Policies?"*

### 31.3 for advanced economies. It also varies between countries, ranging from 24.8 in Slovenia

### Output Losses in Europe during COVID-19: What Role for Policies? — Inequality, Informality, and IMF-supported Programs

### Inequality and fiscal policy effectiveness
- Cross-country variation: "31.3 for advanced economies. It also varies between countries, ranging from 24.8 in Slovenia to 42 in Turkey."
- Theoretical ambiguity:
  - Higher inequality can imply a higher proportion of liquidity-constrained households with a high marginal propensity to consume, increasing fiscal multipliers (citing Brinca et al. (2016)).
  - Conversely, greater inequality concentrated at the top may cause a higher proportion of government transfers to be saved, reducing fiscal multipliers.
- Empirical finding (Panel A, Table 2):
  - "The results from Panel A suggest that higher levels of inequality are associated with a higher fiscal multiplier when non-linearities are accounted for."
  - This is consistent with the hypothesis that economies with a higher proportion of liquidity-constrained households had larger fiscal multipliers during the pandemic.

### Informality and fiscal policy efficacy
- Average informality (average contribution to GDP of the shadow economy, 2015–17):
  - Emerging countries: "29.3 percent of GDP"
  - Advanced economies: "14.8 percent"
- Channels affecting fiscal efficacy:
  - Informality can decrease effectiveness because informal workers may not access furlough, unemployment benefits, or other fiscal supports, thereby reducing fiscal multipliers.
  - Informality can increase fiscal multipliers by limiting enforcement of containment measures that curtail spending.
- Empirical finding (Panel B, Table 2):
  - "The interactive term between informality and fiscal policy support are positive, but significant only in some specifications."
  - These results "suggest that a higher level of informality is weakly associated with a higher efficacy of fiscal policy."
  - Overall implication: findings point towards a relatively higher impact of fiscal policies in emerging countries.

### Regression evidence (Table 2 summary)
- Dependent variable: "GDP growth during 2020 at the country level, where we remove underlying growth and the contribution of the sectoral composition."
- Controls: regressions include country fixed effects and control for Stringency index, de facto mobility, Central Bank policy rate and central bank balance sheet expansion. Robust pval in parentheses. *** p<0.01, ** p<0.05, * p<0.1.
- Selected coefficient highlights:
  - Total fiscal package coefficients (Panel A, columns shown): "0.404***", "0.948***", "0.985***", "0.922***", "0.767***" with p-values "(0.000)".
  - Total fiscal package, quadratic term: negative and significant, e.g., "-0.021***", "-0.023***", "-0.028***", "-0.025***" with p-values "(0.000)".
  - Fiscal package * Inequality (quartiles): "0.325***" with p-value "(0.002)".
  - Panel B total fiscal package coefficients: "0.491***", "0.352***", "0.194", "0.964***", "0.897***", "0.790***" with corresponding p-values.
  - Fiscal package * Informality (level): positive in some specs "0.180*" and "0.107*" (p-values indicating weaker significance).
- Sample and fit:
  - Observations: "123" (across columns)
  - R-squared values reported, e.g., "0.784", "0.842", "0.849", "0.863", "0.853" (Panel A); "0.798", "0.789", "0.847", "0.846" (Panel B).

### Role of IMF-supported programs
- IMF assistance during 2020: "Albania, Moldova, Ukraine, Bosnia and Herzegovina and Kosovo received IMF funding in 2020 through IMF-supported programs."
- Mechanisms by which IMF programs may raise fiscal multipliers:
  - IMF financing may relax fiscal space constraints, helping countries enact stimulus without crowding out private investment or raising sovereign yields.
  - IMF programs may catalyze capital inflows by signaling sound macroeconomic policies.
  - Enhanced policy recommendations, technical assistance, and, in some cases, conditionalities may increase the effectiveness of fiscal policies (e.g., targeting stimulus to higher-multiplier elements).
- Empirical test (Table 3 interaction):
  - Interaction of fiscal policy support with a dummy for IMF-supported program in 2020 yields a significant positive coefficient, "indicating that the fiscal multiplier is higher in countries with IMF programs."
- Important caveats:
  - Possible selection bias: countries with the tightest fiscal space constraints are more likely to request IMF programs and may also have higher returns to marginal fiscal impulse, potentially biasing the interaction coefficient.
  - IMF program recipients may also have received disbursements from other official creditors; positive coefficients could capture broader international support rather than IMF programs alone.

### Differences in marginal fiscal multipliers and decomposition
- Method: compute country-by-country marginal fiscal multiplier using coefficients from regressions and country-specific data on inequality, informality, IMF programs, fiscal policy magnitudes and compositions; set effect to 0 if the quadratic term is larger than the linear term.
- Figure 8 and accompanying analysis:
  - "The figures show larger multipliers for additional fiscal support in emerging countries compared to advanced countries."
  - Explanatory factors for larger marginal multipliers in emerging countries: higher share of above-the-line measures in fiscal packages, smaller fiscal packages, higher inequality, larger shadow economy, and presence of IMF-supported programs.
  - Despite higher estimated multipliers in emerging countries, "the role of policies in advanced economies was, however, much more important in mitigating the crisis than in emerging countries" due to "the considerably larger size of announced policies in advanced economies."
- Calibration/decomposition result:
  - Policy measures raised the potential contribution of policies by "over 70 percent in advanced economies and more than doubling it in emerging economies."

### Conclusion — key takeaways
- Major drivers of cross-country growth differentials during the pandemic: differentials in underlying growth, decline in mobility, pre-pandemic health and macroeconomic fundamentals, and policy support measures; sectoral composition played a limited role.
- Mobility decline in 2020 "contributes the most to output losses in all countries."
- Shallower recessions in emerging countries were due to "higher underlying growth and younger populations which are less at-risk from COVID-19 infections," despite larger policy support in advanced economies.
- Fiscal multiplier determinants: larger where above-the-line measures account for a higher proportion of the fiscal package; where total fiscal packages are smaller; where inequality is higher; where the shadow economy is larger; and in countries with an IMF-supported program in place during the pandemic.
- Policy implication: heterogeneity in country characteristics and policy composition matters materially for the effectiveness of fiscal support during a crisis.

*Source: IMF Working Paper excerpt — "Output Losses in Europe during COVID-19: What Role for Policies?"*

### References

### References

### Key citations (as listed in source)
- Acemoglu, D., Chernozhukov, V., Werning, I. and Whinston, M.D. (2020). A multi-risk SIR model with optimally targeted lockdown (Vol. 2020). Cambridge, MA: National Bureau of Economic Research.  
- Alfaro, L., Chari, A., Greenland, A. N., & Schott, P. K. (2020). Aggregate and firm-level stock returns during pandemics, in real time (No. w26950). National Bureau of Economic Research.  
- Altig, D., Baker, S., Barrero, J.M., Bloom, N., Bunn, P., Chen, S., Davis, S.J., Leather, J., Meyer, B., Mihaylov, E. and Mizen, P. (2020). Economic uncertainty before and during the COVID-19 pandemic. Journal of Public Economics, 191, p.104274.  
- Alvarez, F.E., Argente, D. and Lippi, F. (2020). A simple planning problem for covid-19 lockdown (No. w26981). National Bureau of Economic Research.  
- Andersen, A. L., Hansen, E. T., Johannesen, N., & Sheridan, A. (2020). Consumer responses to the COVID-19 crisis: Evidence from bank account transaction data. Available at SSRN 3609814.  
- Ari, A., Chen, S. and Ratnovski, L. (2021). The dynamics of non-performing loans during banking crises: A new database with post-COVID-19 implications, Journal of Banking & Finance, Elsevier, vol. 133(C).  
- Aum, S., Lee, S.Y.T. and Shin, Y. (2020). Who should work from home during a pandemic? the wage-infection trade-off (No. w27908). National Bureau of Economic Research.  
- Baldwin, R., Weder di Mauro, B. (2020). Mitigating the COVID Economic Crisis: Act Fast and Do Whatever It Takes. VoxEU.org Book, CEPR Press.  
- Baker, S. R., Bloom, N., Davis, S. J., & Terry, S. J. (2020). Covid-induced economic uncertainty (No. w26983). National Bureau of Economic Research.  
- Bartik, A.W., Bertrand, M., Cullen, Z.B., Glaeser, E.L., Luca, M. and Stanton, C.T. (2020). How are small businesses adjusting to COVID-19? Early evidence from a survey (No. w26989). National Bureau of Economic Research.  
- Barro, R.J., Ursúa, J.F. and Weng, J. (2020). The coronavirus and the great influenza pandemic: Lessons from the “spanish flu” for the coronavirus’s potential effects on mortality and economic activity (No. w26866). National Bureau of Economic Research.  
- Bayer, C., B. Born, R. Luetticke, and Müller, G. J. (2020). The Coronavirus Stimulus Package: How large is the transfer multiplier? CEPR Discussion Papers 14600.  
- Becker, T., and Mauro, P. (2006). Output Drops and the Shocks That Matter. IMF Working Papers, (172).  
- Bricco, J.G., Misch, F. and Solovyeva, A. (2020). What are the Economic Effects of Pandemic Containment Policies? Evidence from Sweden.  
- Brinca, P., Holter, H., Krussell, P., Malafry, L. (2016). Fiscal multipliers in the 21st century. Journal of Monetary Economics, Volume 77, February 2016, Pages 63-69.  
- Burriel, P. and Galesi, A. (2018). Uncovering the heterogeneous effects of ECB unconventional monetary policies across euro area countries. European Economic Review 101, 210-229.  
- Büyükbaşaran, T., Can, G.K. and Küçük, H. (2019). Identifying Credit Supply Shocks in Turkey (No. 1906). Central Bank of the Republic of Turkey, Head Office, Structural Economic Research Department.  
- Caceres, C., Gruss, B., and Weber, S. (2021). Differences in Output Performance Between Europe and the US During Covid-19. April 2021 Regional Economic Outlook Update for Europe, International Monetary Fund.  
- Cerra, V., and Saxena, S. C. (2008). Growth dynamics: the myth of economic recovery. American Economic Review, 98(1), 439-57.  
- Cerra, M. V., and Saxena, M. S. C. (2017). Booms, Crises, and Recoveries: A New Paradigm of the Business Cycle and its Policy Implications IMF Working Papers, (No. 2017/250).  
- Chudik, A., Mohaddes, K., and Raissi, M. (2021). Covid-19 fiscal support and its effectiveness” Economics Letters, 205, 109939.  
- Correia, S., Luck, S. and Verner, E., (2020). Pandemics depress the economy, public health interventions do not: Evidence from the 1918 flu. SSRN 2020.  
- Deb, P., Furceri, D., Ostry, J.D. and Tawk, N., (2022). The economic effects of Covid-19 containment measures. Open Economies Review, 33(1), pp.1-32.  
- Eichenbaum, M.S., Rebelo, S. and Trabandt, M. (2021). The macroeconomics of epidemics. The Review of Financial Studies, 34(11), pp.5149-5187.  
- Emmerling, J., Furceri, D., Monteiro, F.L., Pizzuto, P. and Tavoni, M. (2021). Will the economic impact of COVID-19 persist? Prognosis from 21st century pandemics. IMF Working Papers, 2021(119).  
- Faria-e-Castro, M. (2021). Fiscal policy during a pandemic. Journal of Economic Dynamics and Control 125, 104088.  
- Favero, C.A., Ichino, A. and Rustichini, A. (2020). Restarting the economy while saving lives under Covid-19.  
- Fetzer, T., Witte, M., Hensel, L., Jachimowicz, J.M., Haushofer, J., Ivchenko, A., Caria, S., Reutskaja, E., Roth, C., Fiorin, S. and Gomez, M. (2020). Global behaviors and perceptions in the COVID-19 pandemic.  
- Gambacorta, L., Hofmann, B. and Peersman. G. (2014). The effectiveness of unconventional monetary policy at the zero lower bound: A cross‐country analysis. Journal of Money, Credit and Banking 46(4), 615-642.  
- Guerrieri, V., Lorenzoni, G., Straub, L. and Werning, I. (2020). “Macroeconomic implications of COVID-19: Can negative supply shocks cause demand shortages?” American Economic Review, 112(5), 1437-74.  
- Gopinath, G. (2020). Limiting the economic fallout of the coronavirus with large targeted policies. Mitigating the COVID economic crisis: Act fast and do whatever it takes , 41-48.  
- Hanke, M., Kosolapova, M., & Weissensteiner, A. (2020). COVID-19 and market expectations: Evidence from option-implied densities. Economics Letters, 195, 109441.  
- Jarociński, M. (2010). Responses to monetary policy shocks in the east and the west of Europe: a comparison. Journal of Applied Econometrics 25(5), 833-868.  
- Jones, C., Philippon, T. and Venkateswaran, V., (2020). A note on efficient mitigation policies. Covid Economics, 4, pp.25-46.  
- Kermack, W.O. and McKendrick, A.G., (1927). A contribution to the mathematical theory of epidemics. Proceedings of the royal society of london. Series A, Containing papers of a mathematical and physical character, 115(772), pp.700-721.  
- Kılınç, M., and Tunç, C. (2014). Identification of Monetary Policy Shocks in Turkey: A Structural VAR Approach. Working Papers 1423, Research and Monetary Policy Department, Central Bank of the Republic of Turkey.  
- Ma, C., Rogers, J.H. and Zhou, S., (2020). Global economic and financial effects of 21st century pandemics and epidemics. Covid Economics, 5, pp.56-78.  
- Maloney, W.F. and Taskin, T., (2020). Determinants of social distancing and economic activity during COVID-19: A global view. World Bank Policy Research Working Paper, (9242).  
- Medina, L., and Schneider, M. F. (2018). Shadow economies around the world: what did we learn over the last 20 years? International Monetary Fund.  
- Milesi-Ferretti, G.M. (2021). “Travel Shock.” Hutchins Center Working Paper No. 74. The Brookings Institution, Washington, DC.  
- Mountford, A. (2005). Leaning into the wind: a structural VAR investigation of UK monetary policy. Oxford Bulletin of Economics and Statistics 67(5), 597-621.  
- Sheridan, A., Andersen, A. L., Hansen, E. T., & Johannesen, N. (2020). Social distancing laws cause only small losses of economic activity during the COVID-19 pandemic in Scandinavia. Proceedings of the National Academy of Sciences, 117(34), 20468-20473.  
- Vymyatnina, Y. (2005). Monetary policy transmission and Bank of Russia monetary policy. European University at Saint-Petersburg Department of Economics Working Paper.  
- Weale, M. and Wieladek, T. (2016). What are the macroeconomic effects of asset purchases? Journal of Monetary Economics 79, 81-93.

### Appendix — Tables A1–A6 (selected content and key statistics)

- Table A1. Data Sources and Construction — Variables and data sources listed include:
  - Initial conditions: Trade openness (Defined as sum of imports and exports divided by GDP) — IMF WEO.
  - Current account balance — IMF WEO.
  - Gini inequality index — WB WDI.
  - Size of shadow economy (As a share of official GDP in 2017) — Medina & Schneider (2018).
  - Median age — UN Population Division, World Population Prospects, 2017 Revision.
  - Hospital beds per 1000 people — OECD, Eurostat, WB WDI, National Authorities.
  - Share of smokers in population (Average of male and female smokers ratios) — WB WDI.
  - Population density (People per sq. km of land) — WB WDI.
  - Policy: Fiscal support measures (Announced measures as percent of 2019 GDP. Time variation reflects different vintages of the survey) — IMF COVID-19 Policy Survey.
  - Real interest rates (Ex-post real interest rates calculated as key policy rates less CPI inflation, in quarterly averages.) — Haver Analytics, Eurostat, European Central Bank, National Authorities, IMF staff calculations.
  - Central bank assets (As percent of 2019 GDP) — Haver Analytics, European Central Bank, National Authorities.
  - Mobility: Stringency of containment measures (Quarterly average of higher frequency data) — Blavatnik School of Government at the University of Oxford.
  - De facto mobility (Quarterly average of residuals from a weekly panel regression with google mobility ... ) — Google Mobility Reports, IMF staff calculations.

- Table A2. Results for the Sectoral Panel Regressions — Key model summary statistics:
  - Dependent variable: 2020 Real GDP growth net of underlying growth & sectoral composition effects; sectoral columns (1)–(11) correspond to Agriculture; Industry exc. Cons.; Construction; Wholesale & retail; ICT; Finance & insurance; Real estate; Prof., scientific & tech; Pub. adm., educ. & social work; Arts & other.
  - Fiscal support measures coefficients (column values): 0.274*** (col 1), 0.126* (col 2), 0.360*** (col 3), 0.214** (col 4), 0.406*** (col 5), 0.105** (col 6), 0.0653 (col 7), 0.0239 (col 8), 0.230*** (col 9), 0.126*** (col 10), 0.319** (col 11).
  - Stringency of containment coefficients (column values): -0.178***, -0.0825***, -0.140***, -0.168***, -0.338***, -0.102***, -0.0448**, -0.0470***, -0.258***, -0.0720***, -0.485*** (cols 1–11 respectively).
  - De facto mobility coefficients (column values): 0.144**, 0.0660, 0.163, 0.225, 0.148, 0.0943, 0.0481, 0.0779**, 0.0866, 0.105, 0.166* (cols 1–11 respectively).
  - Median age, Population density, and other controls included with reported coefficients and robust standard errors (see table).
  - Observations: 164 (for each column).  
  - R-squared values by column: 0.616 (col 1), 0.123 (col 2), 0.321 (col 3), 0.225 (col 4), 0.594 (col 5), 0.316 (col 6), 0.176 (col 7), 0.234 (col 8), 0.502 (col 9), 0.265 (col 10), 0.460 (col 11).  
  - Note: Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Column (1) is for demonstration only while columns (2)-(11) reflect sectoral regressions used in growth decomposition.

- Table A3. Calibration of Monetary Policy Multipliers — Selected country entries:
  - Table headline: Calibration of Monetary Policy Multipliers. Columns: Impact of 1 p.p. cut in policy rate; Literature reference; Impact of increase in central bank assets by 1% of 2019 GDP; Literature reference.
  - Albania: 0.33; Jarociński (2010); 0.24; Burriel & Galesi (2018).
  - Austria: 0.42; Jarociński (2010); 0.11; Burriel & Galesi (2018).
  - Belgium: 0.42; Jarociński (2010); 0.06; Burriel & Galesi (2018).
  - Estonia: 0.42; Jarociński (2010); 0.33; Burriel & Galesi (2018).
  - Russia: 0.14; Vymyatnina (2005); 0.24; Burriel & Galesi (2018).
  - Turkey: 0.75; Büyükbaşaran, Can & Küçük (2019); 0.24; Burriel & Galesi (2018).
  - United Kingdom: 0.43; Mountford (2005); 0.25; Weale & Wieladek (2016).
  - Note: The calibrated multipliers for an increase in central bank assets in Albania, Belarus, Bosnia and Herzegovina, Bulgaria, Croatia, Czech Republic, Hungary, Kosovo, North Macedonia, Moldova, Montenegro, Poland, Romania, Russia, Serbia, Turkey, Ukraine are extrapolated from the average of the multipliers estimated by Burriel & Galesi (2018) for Estonia, Latvia, Lithuania and the Slovak Republic.

- Table A4. Regression Results on Fiscal Composition — Selected regression outputs:
  - Variables included in columns (1)–(4): ATL; Liquidity; Other; squared terms ATL, Liquidity, Other; Policy rate; CB Balance Sheet; Stringency Index; De facto mobility; Fiscal measures; Fiscal measures, quadratic.
  - ATL coefficients: 1.032*** (col 1), 2.413*** (col 2).  
  - Other coefficient examples: 0.218** (col 1), 0.493*** (col 2).  
  - Stringency Index coefficients: -0.165*** (col 1), -0.104*** (col 2), -0.123*** (col 3), -0.0578* (col 4).  
  - De facto mobility coefficients: 0.151* (col 1), 0.159** (col 2), 0.181** (col 3), 0.148** (col 4).  
  - Fiscal measures coefficients: 0.404*** (col 3), 0.948*** (col 4).  
  - Observations: 123 (all columns). R-squared: 0.784 (col 1), 0.842 (col 2), 0.819 (col 3), 0.865 (col 4).  
  - Robust p-values in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Country FE = YES, Time FE = NO.

- Table A5. Regression Results on Fiscal Multipliers given by Inequality and Informality — Selected highlights:
  - Baseline Fiscal measures coefficient across multiple specifications: 0.404*** (cols 1–3), 0.428*** (col 4), 0.491*** (col 5), 0.352*** (col 6), 0.194 (col 7), 0.948*** (col 8), 0.985*** (col 9), 0.922*** (col 10), 0.767*** (col 11), 0.964*** (col 12), 0.897*** (col 13), 0.790*** (col 14).
  - Fiscal * Informality (level) coefficient examples: 0.180* (col 7), 0.107* (col 8).
  - De facto mobility remains positive and significant in many columns (e.g., 0.151* col 1; 0.159** col 8).  
  - Fiscal, quadratic coefficients: -0.0211*** (col 1), -0.0226*** (col 2), -0.0283*** (col 3), etc.
  - Observations: 123 (all columns). R-squared values range from 0.784 to 0.863 across specifications. Country FE = YES, Time FE = NO.

- Table A6. Regression Results on Fiscal Multipliers and IMF-supported programs — Selected highlights:
  - Fiscal measures coefficients across specifications: 0.404*** (col 1), 0.407*** (col 2), 0.947*** (col 3), 0.410*** (col 4), 0.418*** (col 5), 0.443*** (col 6), 0.486*** (col 7), 0.989*** (col 8), 0.927*** (col 9), 0.777*** (col 10), 0.973*** (col 11).
  - Fiscal * IMF interaction coefficients: 1.624** (col 1), 1.584** (col 2), 1.891** (col 3), 1.174*** (col 4), -0.337 (col 5), 2.457 (col 6), 1.876*** (col 7), 1.032 (col 8), -0.178 (col 9), 2.817* (col 10).
  - Stringency Index negative and significant across specifications (e.g., -0.165*** col 1; -0.135*** col 2).  
  - De facto mobility positive and significant in many specifications (e.g., 0.151* col 1; 0.161** col 2; 0.169*** col 3).  
  - Observations: 123 (all columns). R-squared ranges from 0.784 to 0.876. Country FE = YES, Time FE = NO. Robust p-values in parentheses. *** p<0.01, ** p<0.05, * p<0.1.

*Output Losses in Europe during COVID-19: What Role for Policies? Working Paper No. WP/2022/130*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2022/english/wpiea2022130-print-pdf.pdf_