## Steady but Slow: Resilience amid Divergence

## Source details

**Canonical URL:** [Steady but Slow: Resilience amid Divergence](https://www.imf.org/-/media/files/oap/oap-home/2024/weooutreach-050824-oap-v2.pdf)

## Other formats

- [Markdown version](/-/media/files/oap/oap-home/2024/weooutreach-050824-oap-v2.pdf.md)
- [Structured JSON version](/-/media/files/oap/oap-home/2024/weooutreach-050824-oap-v2.pdf.json)

---

### Overview: Disinflation amid Resilience
- The World Economic Outlook, April 2024 highlights that economic activity has been "surprisingly resilient."
- Risks to the global outlook are described as "broadly balanced."
- Near-term priorities: ensure inflation "touches down smoothly" and rebuild fiscal buffers.
- Data samples and coverage:
  - Median charts based on a sample of 57 economies that accounts for 78 percent of World Economic Outlook world GDP (in weighted purchasing-power-parity terms) in 2023.
  - Real GDP index benchmark: 2020:Q4 = 100 (right-panel sample based on 44 economies).
  - Bands depict the 25th to 75th percentiles across economies.

### Inflation and Output Dynamics
- Global inflation is falling while output grows; headline inflation plotted as month-over-month, SAAR, with series for World, AEs, and EMDEs.
- Headline and core inflation measures:
  - Headline inflation: percent; month-over-month, SAAR.
  - Core inflation: more "sticky" relative to headline (percent, three-month moving average; SAAR).
- Different drivers of inflation across regions include residual, labor market tightness, longer-term expectations, pass-through effects, and headline inflation shocks (IMF staff estimates).

### Surprising Strength since 2022
- Cumulative GDP growth and inflation deviations relative to October 2022 forecasts show surprising strength for several groups:
  - Groups referenced: US, EMxCHN, World, AEs, China, EA, LIDCs.
- Favorable demand developments:
  - Particularly strong private and public consumption relative to projections from October 2022.
  - Accumulated excess savings from the pandemic financed consumption, shown as percent of GDP across economies and over time (2019:Q4–2023:Q4 for selected countries).
- Favorable supply developments:
  - Unclogged supply chains: Global supply chain pressure index and World container index show normalization since pandemic.
  - Increased labor supply: indices of domestic- and foreign-born in the labor force (Index, January 2019 = 100) for US and euro area.

### Medium-Term Growth and Convergence
- "Steady but Too Slow": Five-year-ahead growth projections indicate a longer path to income convergence.
- Per-capita GDP growth over the medium term shown for: BRA, CHN, IND, KOR, RUS, ZAF, USA.
- Projections referenced from April WEO of 2008 and 2023; medium term defined as five-year forecast. Bubble size represents population in year t.

### Risks to the Outlook: Broadly Balanced
- Upside scenarios:
  - Short-term fiscal boost in the context of elections.
  - Further supply-side surprises allowing for faster monetary policy easing.
  - Spurs to productivity from artificial intelligence.
- Downside scenarios:
  - New commodity price spikes amid regional conflicts.
  - "Higher for longer" rates leading to financial stress.
  - China recovery faltering.
  - Disruptive fiscal adjustment and debt stress.
  - Intensification of geoeconomic fragmentation.
- Fragmentation effects:
  - Evidence that separation into blocs (hypothetical blocs described) is underway and could result in large output losses, especially in strategic sectors (e.g., chemical and machinery).
  - Trade growth differences measured as percentage point differences before and after war in Ukraine, with within-bloc and between-bloc changes.

### Policy Priorities: From Fighting Inflation to Rebuilding Buffers
- Delivering a smooth landing:
  - Durably restore price stability.
  - Monitor potentially disruptive capital flows as monetary policy becomes less synchronous (IPF).
- Rebuilding fiscal buffers:
  - Build credibility with sound frameworks.
  - Fiscal consolidation becomes easier as monetary policy eases gradually.
  - Calibrate pace of adjustment to avoid disruptions.
  - Address debt distress.
- Fostering faster productivity growth:
  - Carefully sequence reforms.
  - Harness potential of AI.
- Fiscal consolidation guidance:
  - Many AEs and EMDEs require fiscal consolidation.
  - A credible medium-term consolidation plan is necessary to avoid adverse market reactions.
  - Historical experience: euro area during 2010–15 illustrates costs of forced adjustment.
- Structural balance presentation:
  - Solid line denotes structural balance from April 2024 WEO; dashed lines denote structural balance forecasts from April and October 2010, 2011 and 2012 WEOs.

### Key Statistics and Metrics (preserved values)
- Sample coverage: 57 economies representing 78 percent of WEO world GDP in 2023.
- Real GDP index benchmark: 2020:Q4 = 100.
- Time window for accumulated excess savings: 2019:Q4–2023:Q4.

---

### Feeling the Pinch? Tracing the Effects of Monetary Policy through Housing Markets (Chapter 2 summary)

### Context and Motivation
- Broad-based and sharp tightening of monetary policy followed an extended "low-for-long" and ultra-low COVID rates period.
- House prices and the resilience of aggregate demand vary significantly across countries (cumulative changes since first hike of this cycle).
- Count of monetary policy actions across groups reported (31 AEs and 20 EMs).

### Key Questions
- What are the housing channels of monetary policy transmission?
- How do housing channels transmit differently across mortgage finance and housing market characteristics?
- Have housing channels weakened over time due to shifting mortgage and housing market characteristics?

### Key Findings (IMF staff)
- Cross-country heterogeneity in housing channels of monetary policy transmission. Transmission is stronger where:
  - The regulatory LTV limits are looser.
  - Household debt ratios are higher.
  - The share of fixed-rate mortgages (FRMs) in stock is lower.
  - Regions face housing supply constraints.
  - Regions experience house price overvaluations.
- Housing channels may have changed over time:
  - Weakening where (1) LTV limits have been tightened, (2) the share of FRMs has increased, or (3) populations moved away from high-density areas.
  - Strengthening where (1) household debt increased, or (2) overvaluations increased most in already-overvalued regions.

### Housing Channels Identified
- Cash flow channels through adjustable-rate mortgages and/or ease of refinancing.
- Expectation / risk premia channel.
- Wealth channels through house prices and collateral values.
- Interest rate channel.
- Credit channels: balance sheet channel and bank lending channel reinforce the interest-rate channel.
- Empirical evidence supports cash flow, expectations, wealth, collateral, and interest-rate channels.

### Data, Methods, and Samples
- Data: Unbalanced panel dataset covering 33 advanced and emerging market economies, 1998:Q4 to 2023:Q1; quarterly frequency.
- Methods: Instrumental variable local projections (LP-IV), following Stock and Watson (2018); Jorda, Schularick and Taylor (2015, 2020), augmented with interaction terms to study heterogeneity.
- Instrument: Policy surprise instrument—surprises around monetary policy announcements among professional forecasters, sourced from Bloomberg—used to instrument lagged policy stance variable.
- Estimation intuition: Interaction coefficient β2_h measures the differential effect of monetary policy depending on ex-ante housing characteristics H_c,t-1 (akin to difference-in-differences).

### Mortgage and Housing Market Characteristics (Cross-Country Patterns)
- Mortgage finance characteristics reported across countries:
  - Share of fixed-rate mortgages (FRM) in stock (residual maturity > 12 months).
  - Household debt-to-GDP levels and changes between 2011:Q1 and 2022:Q4.
  - Regulatory maximum LTV limits in 2011:Q1 and 2021:Q4; some countries tightened LTV limits between those dates.
- Findings on interactions:
  - LTV limits and the share of FRMs complement each other in shaping consumption responses to monetary tightening.
  - Model counterfactuals (based on Chen and others 2023) show differential real private consumption responses to a 100 basis point tightening across combinations:
    - Tight LTV + high FRM
    - Tight LTV + low FRM
    - Loose LTV + high FRM
    - Loose LTV + low FRM
  - Figures show quarters of response and magnitudes as percent of steady-state level (horizontal axis numbers represent quarters).

### Heterogeneity by Housing Market Characteristics
- Housing supply constraints and population density:
  - Greater sensitivity of real GDP per capita to a 100 bps tightening in areas with high population density and supply constraints.
- House price overvaluation:
  - Nominal house price responses following a 100 bps tightening are larger in areas that experienced housing booms/overvaluations.
- Key channels remain the same (cash flow, expectations, wealth, interest-rate, credit) but their relative importance varies by mortgage and housing market structure.

### Policy Implications from Chapter 2
- Where transmission is strong:
  - Monitoring housing markets and debt service capacity and being vigilant for early signs of effects are warranted.
- Where transmission is weak:
  - Erring on the side of too much tightening, while generally less costly, may be riskier this time around.
  - Transmission where fixation periods are shorter may suddenly strengthen.
  - High-for-long interest rates may generate financial stability concerns in some areas.
- Borrower-based macroprudential measures:
  - Can “free the hands of monetary policy” allowing monetary policy to focus on aggregate demand management without fear of precipitating a financial crunch.

---

### Transmission Over Time and Slowdown in Global Medium-Term Growth

### Housing channels and monetary policy transmission (synthesis)
- Strength of housing channels varies across countries; analyses indicate housing channels weakened in several countries.
- Monitoring and macroprudential policy are emphasized to manage trade-offs between monetary policy objectives and financial stability.

### Key findings on global medium-term growth
- Dimming growth prospects reflect secular forces.
- The growth of labor, capital and TFP have all declined, but TFP accounts for the bulk of the decline in growth.
- Among other factors, TFP growth has been held back by a rise in capital and labor misallocation between firms.
- Demographic pressures in major economies have weighed on labor growth; weak business investment has stunted capital formation.
- Assuming no major policy change or shocks, by the end of the decade global growth could fall by 1 pp below its pre-pandemic average of 3.8 percent.
- There is hope for policy reforms and emerging technologies to revive growth prospects.

### Drivers of the historical growth decline
- Historical decline primarily reflects a TFP slowdown; all components contributed:
  - Advanced economies (AEs): all components declined since early 2000s.
  - EMDEs: acceleration in early 2000s, followed by slowdown.
- Convergence dynamics explained about 1/4 of the decline from pre-GFC growth rate.
- Demographic drag:
  - Working-age population growth comparisons (2008 vs. 2021) show major economies facing demographic headwinds.
  - Breakdown of change in labor force participation 2008–21 highlights regional differences.
- Capital formation:
  - Business investment held back by reduced growth and firm-specific factors (Tobin’s Q; leverage).
  - Increased capital flows to EMMIEs have supported investment rates.
- Productivity and misallocation:
  - Allocative efficiency (AL) declined for major economies, exerting a drag on TFP growth.
  - Change in AL decomposed using sector-level computations (Hsieh and Klenow (2009); Bils et al. (2021)).
  - Rise of within-sector misallocation accounts for most of the change in allocative efficiency:
    - 70 percent within-sector.
    - 30 percent from sector shares.
  - Contribution of AL to annual TFP growth estimated for a sample of 13 goods and 6 services sectors and 20 economies.

### Quantitative medium-term projections (baseline)
- Medium term (2030) baseline projections:
  - TFP growth: 0.9 percent (2000-19: 1.0 percent).
  - Labor growth: 0.2 percent (2000-19: 0.7 percent).
  - Capital growth: 1.7 percent (2000-19: 2.1 percent).
- As a result, world real GDP growth implies a shortfall of 1 percentage point compared to pre-pandemic average.

### Scenarios and potential policy or shock impacts (relative to baseline, in percentage points)
- Policy interventions:
  - Boosting labor force participation.
  - Migration boost in AEs.
  - Reforms tackling misallocation.
  - Improving talent allocation in EMDEs.
- Emerging technologies:
  - Artificial intelligence (AI) adoption.
- Headwinds:
  - Legacy of high public debt.
  - Geoeconomic fragmentation.
- Projection methodology notes:
  - Five-year-ahead real GDP growth projections use Consensus Economics and IMF staff calculations; WEO sample comprises 196 economies and Consensus Economics sample comprises 88 economies.
  - Example: the 2029 forecast is based on the April 2024 WEO.

### Policy recommendations and reform priorities
- Revive growth prospects through strong policy action and harnessing emerging technologies.
- Structural reforms to reduce misallocation between firms.
- Facilitate merit-based talent allocation, especially in EMDEs.
- Support labor market participation via targeted policies:
  - For males in prime working age: lowering the labor tax wedge and unemployment benefits.
  - For females in prime working age: expansion in secondary education, active labor market programs and childcare support.
  - For older workers: retirement age reforms and active labor market programs.
- Avoid further geoeconomic fragmentation.
- Closing policy gaps in low-AL countries toward the USA could substantially support medium-term growth.
- Approximately one-third of measured misallocation is transitory; two-thirds are structural—policy should address both components.

*Source: World Economic Outlook, April 2024; Office for Asia and the Pacific, May 8, 2024; International Monetary Fund — World Economic Outlook, April 2024 (Chapter 3 and Transmission Over Time).*

### Section 1

### Steady but Slow: Resilience amid Divergence

### Overview: Disinflation amid Resilience
- The World Economic Outlook, April 2024 highlights that economic activity has been "surprisingly resilient."
- Risks to the global outlook are described as "broadly balanced."
- Near-term priorities: ensure inflation "touches down smoothly" and rebuild fiscal buffers.
- Data samples and coverage:
  - Median charts based on a sample of 57 economies that accounts for 78 percent of World Economic Outlook world GDP (in weighted purchasing-power-parity terms) in 2023.
  - Real GDP index benchmark: 2020:Q4 = 100 (right-panel sample based on 44 economies).
  - Bands depict the 25th to 75th percentiles across economies.

### Inflation and Output Dynamics
- Global inflation is falling while output grows; headline inflation plotted as month-over-month, SAAR, with series for World, AEs, and EMDEs.
- Headline and core inflation measures:
  - Headline inflation: percent; month-over-month, SAAR.
  - Core inflation: more "sticky" relative to headline (percent, three-month moving average; SAAR).
- Different drivers of inflation across regions include residual, labor market tightness, longer-term expectations, pass-through effects, and headline inflation shocks (IMF staff estimates).

### Surprising Strength since 2022
- Cumulative GDP growth and inflation deviations relative to October 2022 forecasts show surprising strength for several groups:
  - Groups referenced: US, EMxCHN, World, AEs, China, EA, LIDCs.
- Favorable demand developments:
  - Particularly strong private and public consumption relative to projections from October 2022.
  - Accumulated excess savings from the pandemic financed consumption, shown as percent of GDP across economies and over time (2019:Q4–2023:Q4 for selected countries).
- Favorable supply developments:
  - Unclogged supply chains: Global supply chain pressure index and World container index show normalization since pandemic.
  - Increased labor supply: indices of domestic- and foreign-born in the labor force (Index, January 2019 = 100) for US and euro area.

### Medium-Term Growth and Convergence
- "Steady but Too Slow": Five-year-ahead growth projections indicate a longer path to income convergence.
- Per-capita GDP growth over the medium term shown for: BRA, CHN, IND, KOR, RUS, ZAF, USA.
- Projections referenced from April WEO of 2008 and 2023; medium term defined as five-year forecast. Bubble size represents population in year t.

### Risks to the Outlook: Broadly Balanced
- Upside scenarios:
  - Short-term fiscal boost in the context of elections.
  - Further supply-side surprises allowing for faster monetary policy easing.
  - Spurs to productivity from artificial intelligence.
- Downside scenarios:
  - New commodity price spikes amid regional conflicts.
  - "Higher for longer" rates leading to financial stress.
  - China recovery faltering.
  - Disruptive fiscal adjustment and debt stress.
  - Intensification of geoeconomic fragmentation.
- Fragmentation effects:
  - Evidence that separation into blocs (hypothetical blocs described) is underway and could result in large output losses, especially in strategic sectors (e.g., chemical and machinery).
  - Trade growth differences measured as percentage point differences before and after war in Ukraine, with within-bloc and between-bloc changes.

### Policy Priorities: From Fighting Inflation to Rebuilding Buffers
- Delivering a smooth landing:
  - Durably restore price stability.
  - Monitor potentially disruptive capital flows as monetary policy becomes less synchronous (IPF).
- Rebuilding fiscal buffers:
  - Build credibility with sound frameworks.
  - Fiscal consolidation becomes easier as monetary policy eases gradually.
  - Calibrate pace of adjustment to avoid disruptions.
  - Address debt distress.
- Fostering faster productivity growth:
  - Carefully sequence reforms.
  - Harness potential of AI.
- Fiscal consolidation guidance:
  - Many AEs and EMDEs require fiscal consolidation.
  - A credible medium-term consolidation plan is necessary to avoid adverse market reactions.
  - Historical experience: euro area during 2010–15 illustrates costs of forced adjustment.
- Structural balance presentation:
  - Solid line denotes structural balance from April 2024 WEO; dashed lines denote structural balance forecasts from April and October 2010, 2011 and 2012 WEOs.

---

### Feeling the Pinch? Tracing the Effects of Monetary Policy through Housing Markets (Chapter 2 summary)

### Context and Motivation
- Broad-based and sharp tightening of monetary policy followed an extended "low-for-long" and ultra-low COVID rates period.
- House prices and the resilience of aggregate demand vary significantly across countries (cumulative changes since first hike of this cycle).
- Count of monetary policy actions across groups reported (31 AEs and 20 EMs).

### Key Questions
1. What are the housing channels of monetary policy transmission?
2. How do housing channels transmit differently across mortgage finance and housing market characteristics?
3. Have housing channels weakened over time due to shifting mortgage and housing market characteristics?

### Key Findings (IMF staff)
- Cross-country heterogeneity in housing channels of monetary policy transmission. Transmission is stronger where:
  - The regulatory LTV limits are looser.
  - Household debt ratios are higher.
  - The share of fixed-rate mortgages (FRMs) in stock is lower.
  - Regions face housing supply constraints.
  - Regions experience house price overvaluations.
- Housing channels may have changed over time:
  - Weakening where (1) LTV limits have been tightened, (2) the share of FRMs has increased, or (3) populations moved away from high-density areas.
  - Strengthening where (1) household debt increased, or (2) overvaluations increased most in already-overvalued regions.

### Housing Channels Identified
- Cash flow channels through adjustable-rate mortgages and/or ease of refinancing.
- Expectation / risk premia channel.
- Wealth channels through house prices and collateral values.
- Interest rate channel.
- Credit channels: balance sheet channel and bank lending channel reinforce the interest-rate channel.
- Empirical evidence supports cash flow, expectations, wealth, collateral, and interest-rate channels.

### Data, Methods, and Samples
- Data: Unbalanced panel dataset covering 33 advanced and emerging market economies, 1998:Q4 to 2023:Q1; quarterly frequency.
- Methods: Instrumental variable local projections (LP-IV), following Stock and Watson (2018); Jorda, Schularick and Taylor (2015, 2020), augmented with interaction terms to study heterogeneity.
- Instrument: Policy surprise instrument—surprises around monetary policy announcements among professional forecasters, sourced from Bloomberg—used to instrument lagged policy stance variable.
- Estimation intuition: Interaction coefficient β2_h measures the differential effect of monetary policy depending on ex-ante housing characteristics H_c,t-1 (akin to difference-in-differences).

### Mortgage and Housing Market Characteristics (Cross-Country Patterns)
- Mortgage finance characteristics reported across countries:
  - Share of fixed-rate mortgages (FRM) in stock (residual maturity > 12 months).
  - Household debt-to-GDP levels and changes between 2011:Q1 and 2022:Q4.
  - Regulatory maximum LTV limits in 2011:Q1 and 2021:Q4; some countries tightened LTV limits between those dates.
- Findings on interactions:
  - LTV limits and the share of FRMs complement each other in shaping consumption responses to monetary tightening.
  - Model counterfactuals (based on Chen and others 2023) show differential real private consumption responses to a 100 basis point tightening across combinations:
    - Tight LTV + high FRM
    - Tight LTV + low FRM
    - Loose LTV + high FRM
    - Loose LTV + low FRM
  - Figures show quarters of response and magnitudes as percent of steady-state level (horizontal axis numbers represent quarters).

### Heterogeneity by Housing Market Characteristics
- Housing supply constraints and population density:
  - Greater sensitivity of real GDP per capita to a 100 bps tightening in areas with high population density and supply constraints.
- House price overvaluation:
  - Nominal house price responses following a 100 bps tightening are larger in areas that experienced housing booms/overvaluations.
- Key channels remain the same (cash flow, expectations, wealth, interest-rate, credit) but their relative importance varies by mortgage and housing market structure.

*Source: World Economic Outlook, April 2024; Office for Asia and the Pacific, May 8, 2024; IMF staff calculations and chapter contributions.*

### Section 2

### Transmission Over Time and Slowdown in Global Medium-Term Growth

### Housing channels and monetary policy transmission
- Strength of housing channels varies across countries; analyses indicate housing channels weakened in several countries.
- Where transmission is strong:
  - Monitoring housing markets and debt service capacity and being vigilant for early signs of effects are warranted.
- Where transmission is weak:
  - Erring on the side of too much tightening, while generally less costly, may be riskier this time around.
  - Noteworthy points:
    - Transmission where fixation periods are shorter may suddenly strengthen.
    - High-for-long (interest rates) may generate financial stability concerns in some areas.
- Borrower-based macroprudential measures:
  - Can “free the hands of monetary policy” allowing monetary policy to focus on aggregate demand management without fear of precipitating a financial crunch.

### Key findings on global medium-term growth
- Dimming growth prospects reflect secular forces.
- The growth of labor, capital and TFP have all declined, but TFP accounts for the bulk of the decline in growth.
- Among other factors, TFP growth has been held back by a rise in capital and labor misallocation between firms.
- Demographic pressures in major economies have weighed on labor growth; weak business investment has stunted capital formation.
- Assuming no major policy change or shocks, by the end of the decade global growth could fall by 1 pp below its pre-pandemic average of 3.8 percent.
- There is hope for policy reforms and emerging technologies to revive growth prospects.

### Drivers of the historical growth decline
- Historical decline primarily reflects a TFP slowdown; all components contributed:
  - Advanced economies (AEs): all components declined since early 2000s.
  - EMDEs: acceleration in early 2000s, followed by slowdown.
- Convergence dynamics explained about 1/4 of the decline from pre-GFC growth rate.
- Demographic drag:
  - Working-age population growth comparisons (2008 vs. 2021) show major economies facing demographic headwinds.
  - Breakdown of change in labor force participation 2008–21 highlights regional differences.
- Capital formation:
  - Business investment held back by reduced growth and firm-specific factors (Tobin’s Q; leverage).
  - Increased capital flows to EMMIEs have supported investment rates.
- Productivity and misallocation:
  - Allocative efficiency (AL) declined for major economies, exerting a drag on TFP growth.
  - Change in AL decomposed using sector-level computations (Hsieh and Klenow (2009); Bils et al. (2021)).
  - Rise of within-sector misallocation accounts for most of the change in allocative efficiency:
    - 70 percent within-sector.
    - 30 percent from sector shares.
  - Contribution of AL to annual TFP growth estimated for a sample of 13 goods and 6 services sectors and 20 economies.

### Quantitative medium-term projections (baseline)
- Medium term (2030) baseline projections:
  - TFP growth: 0.9 percent (2000-19: 1.0 percent).
  - Labor growth: 0.2 percent (2000-19: 0.7 percent).
  - Capital growth: 1.7 percent (2000-19: 2.1 percent).
- As a result, world real GDP growth implies a shortfall of 1 percentage point compared to pre-pandemic average.

### Scenarios and potential policy or shock impacts
- Scenario types considered (impact reported relative to baseline, in percentage points):
  - Policy interventions:
    - Boosting labor force participation.
    - Migration boost in AEs.
    - Reforms tackling misallocation.
    - Improving talent allocation in EMDEs.
  - Emerging technologies:
    - Artificial intelligence (AI) adoption.
  - Headwinds:
    - Legacy of high public debt.
    - Geoeconomic fragmentation.
- Projection methodology notes:
  - Five-year-ahead real GDP growth projections use Consensus Economics and IMF staff calculations; WEO sample comprises 196 economies and Consensus Economics sample comprises 88 economies.
  - Example: the 2029 forecast is based on the April 2024 WEO.

### Policy recommendations and reform priorities
- Revive growth prospects through strong policy action and harnessing emerging technologies.
- Structural reforms to reduce misallocation between firms.
- Facilitate merit-based talent allocation, especially in EMDEs.
- Support labor market participation via targeted policies:
  - For males in prime working age: lowering the labor tax wedge and unemployment benefits.
  - For females in prime working age: expansion in secondary education, active labor market programs and childcare support.
  - For older workers: retirement age reforms and active labor market programs.
- Avoid further geoeconomic fragmentation.
- Closing policy gaps in low-AL countries toward the USA could substantially support medium-term growth.
- Approximately one-third of measured misallocation is transitory; two-thirds are structural—policy should address both components.

---


_Source: https://www.imf.org/-/media/files/oap/oap-home/2024/weooutreach-050824-oap-v2.pdf_