## wp1786

## Source details

**Canonical URL:** [wp1786](https://www.imf.org/-/media/files/publications/wp/2017/wp1786.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2017/wp1786.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2017/wp1786.pdf.json)

---

### Introduction and scope
- Definition and scope:
  - Fiscal crises denote periods of extreme fiscal distress when governments are unable to contain large fiscal imbalances, potentially leading to extreme measures such as debt default or monetization of the deficit.
  - Term originated with James O'Connor (1973).
  - Triggers include imbalances or exogenous shocks: financial crises (direct bailouts or indirect through falling tax revenues), loss of market confidence, defaulting on debts, or printing money to finance deficits.
- Research questions:
  - Frequency and duration of crisis episodes.
  - Dynamics before, during, and after crises.
  - Disruptiveness and persistent impacts on economic growth.
  - How to avoid fiscal crises via well-designed fiscal adjustments.
- Contribution:
  - New database of fiscal crises using expanded criteria (including implicit defaults and loss of market access) building on Baldacci et al. (2011).
  - Country coverage: 188 countries; sample period: 1970-2015.
  - Identified fiscal crisis episodes: 436 fiscal crisis episodes; average crises per country in the sample period: two crises between 1970 and 2015.

### Identification methodology (four complementary criteria)
- A year is classified as a fiscal crisis year if at least one of four criteria is met.
- Criterion (1) — Credit events:
  - Any operation that makes creditors incur material economic losses on sovereign debt (default, restructuring, rescheduling).
  - Uses Bank of Canada (BoC, 2016) annual data on nominal stock of sovereign debt obligations in default.
  - Excludes default amounts below 0.2 percent of GDP and continued reporting that grows by less than 10 percent per year.
- Criterion (2) — Exceptionally large official financing:
  - Any year under an IMF financial arrangement with access above 100 percent of quota and fiscal adjustment as a program objective.
  - Excludes precautionary agreements; includes arrangements that became active with access above the threshold.
- Criterion (3) — Implicit domestic public debt default:
  - (3a) Very high inflation thresholds:
    - AMs: ≥ 35 percent per year
    - SDSs: ≥ 35 percent per year
    - EMs: ≥ 100 percent per year
    - LIDCs: ≥ 100 percent per year
  - (3b) Domestic arrears accumulation:
    - Proxy: “other account payables” (OAP)-to-GDP ratio must grow more than 1 percentage point per year.
- Criterion (4) — Loss of market confidence:
  - (4a) Loss of market access: LMA indicator (requires two consecutive years of market access and maintained for at least one fourth of the available sample time).
  - (4b) Price of market access: sovereign spreads threshold at ≥ 1,000 basis points; supplemental EMBI spreads or 10-year local-currency yield spreads and CDS spreads used.

### Key sample statistics and episode counts
- Country coverage: 188 countries; sample period: 1970-2015.
- Total identified fiscal crisis episodes: 436.
- Average number of crises per country since 1970: 2.3.
- Crises by country group (start date within sample period):
  - Total: 436
  - AM: 25
  - EM: 154
  - LIDC: 171
  - SDS: 86
  - Average per country: Total 2.3; AM 0.7; EM 2.2; LIDC 3.4; SDS 2.6
- Episodes with start and end date within sample period:
  - Total: 400
  - AM: 23
  - EM: 143
  - LIDC: 154
  - SDS: 80
  - Average per country: Total 2.1; AM 0.7; EM 2.0; LIDC 3.1; SDS 2.4
- Memorandum items:
  - Ongoing at sample period start: 32 (AM 1; EM 22; LIDC 9; SDS 0)
  - Ongoing at sample period end: 36 (AM 2; EM 11; LIDC 17; SDS 6)
  - Number of countries with no fiscal crisis: 37 (AM 20; EM 9; LIDC 1; SDS 7)
- Comparison with prior studies:
  - This database events: 436
  - Bruns and Poghosyan (2016) update of Baldacci et al. (2011): 201
  - Laeven and Valencia (2012) debt crises: 67
  - Reinhart and Rogoff (2011b) external debt crises: 75
  - Reinhart and Rogoff (2011b) domestic debt crises: 26
  - Number of common events with Bruns and Poghosyan: 141

### Frequency, duration, and temporal patterns
- Crisis years represent almost one third of all observations.
- Average crisis duration: almost 6 years, varying by development group:
  - LIDCs: 6.6 years
  - EMs: 6 years
  - AMs: 3.8 years
  - SDSs: 3.2 years
- Thirty countries have been in crisis more than half of the sample period (majority LIDCs; some EMs).
- Most turbulent decade: the 1990s; the 1980s also saw a large rise in crises, especially among EMs.
- Credit events are the most frequent identification criterion across decades, though less dominant in the latest years.
- Exceptional high-access IMF financing spiked between 2006 and 2010.

### Trigger overlap and twin/triple crises
- Overlap patterns:
  - At least two criteria overlap more than one quarter of the time for crisis years.
  - Only about 9 percent of fiscal crisis episodes start with more than one fulfilled criterion.
  - Most frequent combination: credit events and high-access IMF programs (occurs for more than a fifth of crisis years that saw a credit event).
  - EMs show the highest rates of overlap of different criteria among non-AMs: in 17 crises their start years are identified by two or more criteria.
  - Crises in AMs and EMs show two or more active criteria around one third of the time, compared to a quarter in LIDCs.
  - Economic growth was lower in crisis periods when more than one criterion was met.
- Twin and triple crises:
  - Close to a fifth of fiscal crises coincide with either a banking or currency crisis; 3 percent coincide with both.
  - Vast majority of fiscal-banking crises occurred in AMs and EMs.
  - More than half of fiscal-currency crises occurred in EMs.
  - Triple crises: AMs experienced 3; EMs experienced 11.
  - In 6 triple-crisis cases the fiscal crisis was signaled by more than one criterion; in 5 cases these are among the worst crises in terms of economic growth once the crisis started.

### Economic outcomes around fiscal crises (aggregate stylized facts)
- Crisis windows: pre-crisis t-3 to t-1 and crisis t to t+2.
- Aggregate averages (average, in percent of GDP, except real GDP p.c. growth in percentage points):

  - Full Sample
    - Pre-Crisis* Overall Fiscal Balance -4.0
    - Crisis** Overall Fiscal Balance -6.3
    - Pre-Crisis* Current Account Balance -5.6
    - Crisis** Current Account Balance -5.6
    - Pre-Crisis* Real GDP p.c. growth 2.2
    - Crisis** Real GDP p.c. growth 0.4

  - AM
    - Pre-Crisis* Overall Fiscal Balance -1.9
    - Crisis** Overall Fiscal Balance -3.6
    - Pre-Crisis* Current Account Balance -6.5
    - Crisis** Current Account Balance -2.5
    - Pre-Crisis* Real GDP p.c. growth 2.7
    - Crisis** Real GDP p.c. growth -1.6

  - EM
    - Pre-Crisis* Overall Fiscal Balance -5.0
    - Crisis** Overall Fiscal Balance -9.1
    - Pre-Crisis* Current Account Balance -3.0
    - Crisis** Current Account Balance -4.3
    - Pre-Crisis* Real GDP p.c. growth 2.3
    - Crisis** Real GDP p.c. growth 0.0

  - LIDC
    - Pre-Crisis* Overall Fiscal Balance -3.9
    - Crisis** Overall Fiscal Balance -5.1
    - Pre-Crisis* Current Account Balance -5.8
    - Crisis** Current Account Balance -6.2
    - Pre-Crisis* Real GDP p.c. growth 1.6
    - Crisis** Real GDP p.c. growth 1.1

  - SDS
    - Pre-Crisis* Overall Fiscal Balance -2.6
    - Crisis** Overall Fiscal Balance -3.5
    - Pre-Crisis* Current Account Balance -11.5
    - Crisis** Current Account Balance -9.1
    - Pre-Crisis* Real GDP p.c. growth 2.4
    - Crisis** Real GDP p.c. growth 1.4

  - Note: * t-3 to t-1. ** t to t+2.

- Additional stylized outcomes:
  - Economic growth and the fiscal deficit deteriorate substantially with crisis onset; the current account deficit remains large but stable.
  - AMs suffer the largest growth decline; pre-crisis years show large external imbalances which improve afterwards.
  - EMs experience the largest increases in fiscal deficits around onset and the largest deterioration in current account balance.
  - LIDCs have the smallest decline in economic growth, albeit starting from the lowest pre-crisis level.

### Dynamics around crises — event-study and panel findings
- Method: fixed-effects panel regressions with an 11-year event window (j = -5 to +5), following Gourinchas and Obstfeld (2012).
- Main empirical findings:
  - Fiscal crises tend to be preceded by loose fiscal policy: robust real expenditure growth in the run-up.
  - Once crisis begins, governments aggressively contain expenditure growth — fiscal policy is procyclical as economic conditions weaken.
  - Budget revenues grow at a weaker pace and the fiscal balance continues to worsen during crisis years.
  - Expenditure policy is more procyclical in AMs and EMs and less so in LIDCs around crises.
  - Primary balance deterioration in crisis years is strongest among EMs despite significant reduction in expenditure growth.
  - AMs and SDSs face a more temporary deterioration in the primary balance and undo it faster due to a large upfront deceleration in expenditure growth.
  - Crises signaled by credit events and implicit defaults have the largest increases in the post-crisis budget deficit.
  - Economic growth is generally higher than normal in the run-up, then declines sharply at crisis onset; growth tends to slow in the year preceding the crisis and fall significantly with onset.
  - Fiscal crises triggered by implicit default come with more severe economic deterioration than all others.
  - Growth decline for credit events is mainly driven by cases involving private creditors; when only public creditors are involved, the decline is much smoother and growth converges to normal times.
  - AMs experience the largest fall in real growth in the first two years of the crisis; EMs have the second largest fall (between 3 and 6 percentage points in the first two years).
  - Almost half of AMs and EMs experience negative growth in the first and second year of the crisis; about a third of LIDCs face negative growth in the first year.
  - Public debt levels remain higher than normal throughout the crisis period; at onset public debt ratios rise substantially, especially in AMs and EMs, and fall only gradually afterward.
  - In LIDCs, debt falls with the start of the crisis (from above-average levels), often reflecting debt defaults.
  - Crises where official creditors dominate show pronounced falls in public debt; events involving private creditors tend to keep public debt elevated.
  - Crises identified by IMF program access or loss of market confidence are usually accompanied by larger external imbalances; IMF programs involve substantial upfront cuts to public expenditure growth and a large improvement in the external current account.
  - Twin deficits worsen in the crisis run-up, dominating in AMs and EMs; post-crisis current account adjustment is largest in AMs.
  - Except for SDSs, countries tend to experience a sharp exchange rate depreciation peaking in the year after the start of the crisis.

### Box 1 — Recent wave of fiscal crises (post-GFC features)
- Incidence and timing:
  - Worldwide, 80 fiscal crises occurred after the onset of the global financial crisis (GFC) in 2008.
  - All country groups were affected to a high degree (except SDSs); many recent fiscal crises occurred between 2008 and 2014, with a concentration for AMs between 2008 and 2012.
- AMs:
  - A fifth of AMs went into a fiscal crisis — all between 2008 and 2012.
  - Crisis duration: on average somewhat shorter than previous crises, but stronger in impact.
  - Public debt levels rise, peaking three years after onset; peak was more than 50 percentage points of GDP higher than previous fiscal crises.
  - Crisis triggered when expenditure growth was stronger than past experience; preceded by much larger deterioration in external current account.
  - Economic growth deteriorated significantly only after a few years—by more than 4 percentage points less at its lowest point.
  - Inflation dropped well below previous crises’ levels at crisis onset.
  - Exchange rates experienced a short-lived sizable stronger depreciation shortly after crisis onset, partly fueled by quantitative easing (QE).
- EMs:
  - Two-fifths of EMs went into a fiscal crisis between 2008 and 2014.
  - Affected EMs were mostly financially interconnected Emerging Europe and resource-rich Latin America and Middle East.
  - Pre-crisis: strong growth, lower public debt, smaller fiscal deficits than previous episodes.
  - At crisis onset, debt had converged to average levels and outpaced them a few years after onset (less than AMs).
  - Short-lived large dip in effective interest rates and exchange rate depreciation in first year after onset.
  - Economic growth tended to remain somewhat above levels of previous crises.
- LIDCs:
  - Two-thirds of LIDCs went into fiscal crisis between 2008 and 2014.
  - Started from a stronger position: much lower debt levels (partly reflecting debt relief) and higher growth.
  - Despite acceleration in expenditure growth when the last crisis hit, LIDCs contained fiscal deterioration, partly reflecting stronger real GDP growth; debt levels remained well below those of previous crisis episodes.
- Twin fiscal-financial crises (amplification):
  - Fiscal-banking crises: debt ratios increase by an additional 10 percentage points of GDP; growth is lower by an additional 2½ percentage points on average in the year after onset (amplification especially strong in AMs).
  - Fiscal-currency crises: debt rises by around an additional 10 percent of GDP relative to a fiscal crisis; additional growth contraction about 2 to 3.5 percentage points in the first two years (driven by EMs and some AMs).
- Long-term impacts:
  - Public debt: on average returns to similar level as pre-crisis trend but with group differences:
    - AMs: public debt ratio rises by around 10 percentage points at onset and remains about 4 percentage points above pre-crisis level in the long term.
    - EMs: rise in public debt but eventually converge to pre-crisis level.
    - LIDCs and SDSs: public debt first falls but eventually returns to levels similar to pre-crisis periods.
  - Real GDP p.c.: fiscal crises associated with a permanent loss of real GDP of around 2 percent on average.
    - AMs and EMs: greater output losses but tend to recover half of initial output losses.
    - LIDCs and SDSs: more gradual fall in GDP with no recovery, converging to pre-crisis growth rates over time.

### Data sources, coverage, and empirical notes
- Coverage: 188 countries (all IMF members as of end-2015); 46 years (1970-2015).
- Country groups:
  - AMs: 35
  - EMs: 70
  - LIDCs: 50
  - SDSs: 33
- Primary data sources:
  - Most variables: WEO and previous vintages.
  - Gross debt expanded using FAD’s Historical Public Debt Database with WEO.
  - Sovereign default / credit event: Bank of Canada CRAG sovereign defaults database (BoC (2016)) covering 136 countries from 1970-2015.
  - IMF financial program data: IMF database covering all members from 1952 to 2015.
  - OAP (proxy for arrears): Eurostat and OECD (Eurostat prioritized).
  - Yields: IFS; EMBI: Reuters Datastream; CDS Spreads: Bloomberg.
  - Loss of market access dummies: Guscina, Sheheryar, and Papaioannou (2016) for 57 countries from 1990 onwards; Gelos, Sahay, and Sandleris (2004) for prior years/missing observations.
  - Currency and banking crisis indicators: Laeven and Valencia (2012).
- Key dataset counts:
  - Total observations (Appendix Table 3): 8648.
  - Crisis years (overall reported under Criterion 1 in Appendix Table 3): 1715.
  - Total crisis episodes (Appendix Table 5): 436.
- Percent of episodes with negative real GDP p.c. growth (Appendix Table 5, by timing t-3 to t+2):
  - Total: 24.8 (t-3), 25.5 (t-2), 28.2 (t-1), 40.6 (t), 33.9 (t+1), 29.8 (t+2)
  - AM (25 episodes): 12.0 (t-3), 28.0 (t-2), 28.0 (t-1), 48.0 (t), 60.0 (t+1), 48.0 (t+2)
  - EM (154 episodes): 22.1 (t-3), 20.8 (t-2), 28.6 (t-1), 49.4 (t), 40.3 (t+1), 29.2 (t+2)
  - LIDC (171 episodes): 29.8 (t-3), 28.1 (t-2), 28.1 (t-1), 33.9 (t), 29.2 (t+1), 29.8 (t+2)
  - SDS (86 episodes): 23.3 (t-3), 27.9 (t-2), 27.9 (t-1), 36.0 (t), 24.4 (t+1), 25.6 (t+2)
- Event-study and empirical approaches:
  - Event study window: 11-year window (t-5 to t+5), comparisons to “normal” period average.
  - One outlier dropped: Zimbabwe 2000 (primary balance ratio reduced by more than 1600 percentage points 4 years after onset).
  - Variables analyzed: public debt ratio; effective interest rate; exchange rate depreciation; current account balance; revenue growth; expenditure growth; primary balance; real p.c. GDP growth; inflation rate.
  - Twin-crisis analyses define banking/currency crises occurring in [t-1, t+1] relative to fiscal crisis start year t.
  - Long-term impact analysis uses IRF approach (fixed-effect AR(p) with p=1, q=4); IRFs from 1,000 Monte Carlo simulations with one-standard-error bands.

### Policy-relevant observations and implications
- Crises typically require disruptive fiscal adjustments even when exceptional measures (default, monetization) are taken — such measures do not eliminate need for fiscal consolidation.
- Rapid public debt accumulation during the first three years of crises underscores importance of early preventive measures and monitoring of debt dynamics.
- International official support (e.g., IMF high-access programs) plays a significant role in crisis identification and management; exceptional access often coincides with fiscal adjustment objectives.
- Longer average duration of crises in LIDCs and EMs, and high incidence of credit events affecting both official and private creditors, point to need for coordinated creditor frameworks to manage external and domestic sovereign distress.
- Deeper output declines in twin crises and in AMs/EMs suggest macro-financial linkages and pre-crisis imbalances matter for contraction depth and warrant targeted financial-sector and external-position buffers.
- Suggested future research directions: better identification of economic and policy precursors to crises, development of early warning indicators, and improvement of policy options for prevention and management.

*Source: IMF working paper unit wp1786 (References, Appendix material, and selected chapters excerpt).*

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

### References

### Introduction
- Definition and scope:
  - Fiscal crises denote periods of extreme fiscal distress when governments are unable to contain large fiscal imbalances, potentially leading to extreme measures such as debt default or monetization of the deficit. The term originated with James O'Connor (1973).
  - Fiscal crises can be triggered by imbalances in the economy or exogenous shocks, including financial crises that put budgetary pressure directly (bailouts) or indirectly (sharp economic deterioration and falling tax revenues), loss of market confidence, defaulting on debts, or printing money to finance deficits.

- Research questions:
  - Frequency and duration of crisis episodes.
  - Dynamics around crises (what happens before, during, and after).
  - Disruptiveness and persistent impacts on economic growth.
  - How to avoid fiscal crises, including via well-designed fiscal adjustments.

- Prior literature and gaps:
  - Literature has focused predominantly on sovereign debt defaults and external default episodes (examples cited: Detragiache and Spilimbergo, 2001; Chakrabarti and Zeaiter, 2014).
  - Complementary evidence includes large-scale official financing (Manasse, Roubini, and Schimmelpfennig, 2003) and, to a limited degree, domestic public default (Reinhart and Rogoff, 2009 and 2011; Gourinchas and Obstfeld, 2012).
  - Defaults are important but do not capture all fiscal crises; empirical research on broader fiscal crises is limited.

- Contribution of the paper:
  - Develops a new database of fiscal crises using an expanded set of criteria (including implicit defaults or loss of market access) building on Baldacci et al. (2011).
  - Expands country coverage and sample period to include developing countries underrepresented in the literature.

- Key sample and dataset statistics:
  - Country coverage: 188 countries.
  - Sample period: 1970-2015.
  - Identified fiscal crisis episodes: 436 fiscal crisis episodes.
  - Average crises per country in the sample period: two crises between 1970 and 2015.
  - Empirical evidence includes calculations based on IMFWEO data and the authors' newly created fiscal crises database and highlights episodes with large and abrupt declines in growth (Figure 1.1 references decline in the real GDP p.c. growth rate, percentage points).

### Identification approach and scope of analysis
- Identification focus:
  - Periods of extreme fiscal distress characterized by inability to contain large fiscal imbalances and adoption of extreme measures (explicit examples: debt default, monetization; implicit criteria: implicit defaults, loss of market access).
  - Inclusion of developing countries to capture unique crisis characteristics and broaden empirical coverage.

- Intended empirical analysis (as outlined):
  - Frequency and duration analysis of crisis episodes.
  - Event-study-style analysis of macro-fiscal dynamics before, during, and after crises (implied by references to figures such as Real GDP p.c. growth—Before and During the Crisis and Event Study—Key Macro-Fiscal Variables).
  - Examination of overlaps between fiscal, banking, and currency crises and disaggregation by country groups and crisis criteria.

### Illustrative findings referenced in the text
- Growth impact:
  - Fiscal crises are associated with large and abrupt declines in growth; Figure 1.1 reports the difference in simple averages of real GDP p.c. growth rate of the three years preceding a crisis and the three years after the onset of the crisis, including the onset year (authors' calculations based on IMFWEO data and the newly created fiscal crises database).

### Data and appendices (contents referenced)
- The study leverages a newly created fiscal crises database and IMFWEO data.
- Appendices and annexes include data sources, sample country listings, counts of fiscal crisis years, onset confluence of criteria, overlap between fulfilled criteria, and various appendix figures presenting event-study results on public debt, growth, prices, and key fiscal indicators across country groups and crisis criteria.

*Source: IMF working paper (wp1786) — References and Appendix material excerpt.*

### 2015. While we would expect low-income developing countries (LIDCs) to suffer the most from a

### wp1786 - 2015. While we would expect low-income developing countries (LIDCs) to suffer the most from a

### Overview
- The paper constructs a new database of fiscal crisis episodes covering 188 IMF member countries between 1970 and 2015 to study patterns of fiscal crises and their macro-fiscal dynamics.
- A fiscal crisis is defined as a period of heightened budgetary distress resulting in the sovereign taking exceptional measures (e.g., default, monetization of the deficit).
- The debt dynamics are summarized by the debt-to-GDP identity used in the text:
  - d_t = d_{t−1} * (r_t − g_t)/(1+g_t) − p_t + SF_t
  - where d_t is the debt-to-GDP ratio, r_t is the effective interest rate, g_t is nominal GDP growth, p_t is the primary balance-to-GDP ratio, and SF_t denotes stock-flow adjustments.

### Identification methodology (four complementary criteria)
- A year is classified as a fiscal crisis year when at least one of the four criteria is met.
- Criterion (1) — Credit events
  - Any operation that makes creditors incur material economic losses on sovereign debt (e.g., default, restructuring, or rescheduling).
  - Uses Bank of Canada (BoC, 2016) annual data on nominal stock of sovereign debt obligations in default.
  - Excludes small-scale technical defaults: default amounts below 0.2 percent of GDP.
  - Excludes continued reporting of previously defaulted amounts that grow by less than 10 percent per year.
- Criterion (2) — Exceptionally large official financing
  - Any year under an IMF financial arrangement with access above 100 percent of quota and fiscal adjustment as a program objective.
  - Excludes precautionary agreements; includes arrangements that became active with access above the threshold.
- Criterion (3) — Implicit domestic public debt default
  - Captures (a) very high inflation and/or (b) accumulation of domestic arrears.
  - (3a) Very high inflation thresholds:
    - AMs: ≥ 35 percent per year
    - SDSs: ≥ 35 percent per year
    - EMs: ≥ 100 percent per year
    - LIDCs: ≥ 100 percent per year
  - (3b) Domestic arrears accumulation:
    - Proxy: “other account payables” (OAP)-to-GDP ratio must grow more than 1 percentage point per year.
- Criterion (4) — Loss of market confidence
  - (4a) Loss of market access: identified using an LMA indicator (requires the country to have enjoyed two consecutive years of market access and maintained it for at least one fourth of the available sample time).
  - (4b) Price of market access: sovereign spreads threshold at ≥ 1,000 basis points (bps); supplemental use of EMBI spreads or 10-year local-currency yield spreads and CDS spreads.

### Key findings and stylized facts
- Total identified fiscal crisis episodes: 436.
- Average number of crises per country since 1970: 2.3.
- Crises by country group (with start date within sample period):
  - Total: 436
  - AM: 25
  - EM: 154
  - LIDC: 171
  - SDS: 86
  - Average per country: Total 2.3; AM 0.7; EM 2.2; LIDC 3.4; SDS 2.6
- Episodes with start and end date within sample period:
  - Total: 400
  - AM: 23
  - EM: 143
  - LIDC: 154
  - SDS: 80
  - Average per country: Total 2.1; AM 0.7; EM 2.0; LIDC 3.1; SDS 2.4
- Memorandum items:
  - Ongoing at sample period start: 32 (AM 1; EM 22; LIDC 9; SDS 0)
  - Ongoing at sample period end: 36 (AM 2; EM 11; LIDC 17; SDS 6)
  - Number of countries with no fiscal crisis: 37 (AM 20; EM 9; LIDC 1; SDS 7)
- Comparison with prior studies:
  - Number of events in this database: 436
  - Bruns and Poghosyan (2016) update of Baldacci et al. (2011): 201
  - Laeven and Valencia's (2012) debt crises: 67
  - Reinhart and Rogoff's (2011b) external debt crises: 75
  - Reinhart and Rogoff's (2011b) domestic debt crises: 26
  - Number of common events with Bruns and Poghosyan: 141
- Frequency and duration
  - Crisis years represent almost one third of all observations.
  - Average crisis duration: almost 6 years, varying by development group.
    - LIDCs: 6.6 years
    - EMs: 6 years
    - AMs: 3.8 years
    - SDSs: 3.2 years
  - Thirty countries have been in crisis more than half of the sample period (majority LIDCs; some EMs).
- Temporal patterns
  - Most turbulent decade: the 1990s; the 1980s also saw a large rise in crises, especially among EMs.
  - Credit events are the most frequent identification criterion across decades, though less dominant in the latest years.
  - Exceptional high-access IMF financing spiked between 2006 and 2010 (global economic and financial crisis).
- Triggers and creditor involvement
  - The majority of crises in non-AMs are associated with credit events.
  - For non-AM credit events, about 90 percent involved official creditors and about two-thirds included private creditors.
  - Credit events never signal the start of a crisis for AMs in this sample.
- Economic impact patterns (summary findings)
  - Crises are accompanied by disruptive fiscal adjustments: countries curtail spending growth while economic conditions worsen.
  - Taking exceptional measures (e.g., debt default, monetization of the deficit) does not eliminate the need for a fiscal adjustment.
  - Public debt tends to rise in the first three years of the crisis and only gradually decline afterwards; when crises involve credit events that only involve official creditors, debt falls quickly.
  - Economic growth usually declines sharply at crisis onset.
    - AMs and EMs face the deepest contractions in growth rate (between 3 and 6 percentage points in the first two years of the crisis).
    - About half of AMs and EMs go through recessions in crises.
  - Twin (fiscal and external) crises experience a deeper decline in growth than stand-alone fiscal crises.
  - Evidence from impulse response functions (IRFs) suggests a negative permanent impact on real GDP across all countries; impact on public debt varies.

### Policy-relevant observations and implications (as presented in the text)
- Crises typically require disruptive fiscal adjustments even when exceptional measures are taken, indicating limited short-term relief from actions like default or monetization.
- International official support (e.g., IMF high-access programs) plays a significant role in identifying crises and in crisis management; exceptional access often coincides with fiscal adjustment objectives.
- Rapid accumulation of public debt during the first three years of crises underscores the importance of early preventive measures and monitoring of debt dynamics.
- The longer average duration of crises in LIDCs and EMs, and the high incidence of credit events affecting both official and private creditors, highlight the need for coordinated creditor frameworks and mechanisms to manage both external and domestic sovereign distress.
- The deeper output declines in twin crises and in AMs/EMs suggest that macro-financial linkages and pre-crisis imbalances matter for the depth of contractions and may warrant targeted financial-sector and external-position buffers.

*Source: wp1786 (IMF working paper content unit provided).*

### Appendix Table 2 on crisis years).

### Appendix Table 2 on crisis years)

### Triggering criteria and overlap
- Several identification criteria tend to overlap during the crisis period.  
- When looking at crisis years, at least two criteria overlap more than one quarter of the time (Appendix Table 4).  
- Only about 9 percent of fiscal crisis episodes start with more than one fulfilled criterion (Appendix Table 3).  
- The most frequent combination is credit events and high-access IMF programs, which happens for more than a fifth of the crisis years that saw a credit event.  
- Amongst non-AMs, EMs show the highest rates of overlap of different criteria in the sample. In 17 crises, their start years are identified by two or more criteria (Appendix Table 3).  
- Crises in AMs and EMs show two or more active criteria around one third of the time, compared to a quarter of the time in LIDCs.  
- The data suggests that economic growth was lower in crisis periods when more than one criterion was met (Figure 2.3).

- Trigger counts (as presented in source):
  - AMEMLIDCSDS
  - Credit event08514171
  - Official financing1140296
  - Implicit default131897
  - Market confidence72543

### Fiscal and financial crisis overlaps (twin and triple crises)
- The database contains a number of twin and triple crises (Figure 2.4).  
- Close to a fifth of fiscal crises happen at the same time as either a banking or currency crisis, three percent of them even coincide with both.  
- The vast majority of fiscal-banking crises occurred in AMs and EMs.  
- More than half of the fiscal-currency crises occurred in EMs.  
- Triple crises: AMs experienced three of those triple crises, EMs experienced the other 11.  
- In 6 of the triple-crisis cases, the fiscal crisis was signaled by more than one criterion; in 5 cases they turn out to be some of the worst crises in terms of economic growth once the crisis started.

- Overlap counts (as presented in source):
  - Fiscal crises 352
  - Currency crises 122
  - Banking crises 82
  - 23 
  - 14 
  - 47 
  - 28

### Economic outcomes around fiscal crises (summary of stylized facts)
- Crisis windows considered: pre-crisis t-3 to t-1 and crisis t to t+2.  
- Aggregate averages (Table 3.1; average, in percent of GDP, except real GDP p.c. growth in percentage points):

  - Full Sample
    - Pre-Crisis* Overall Fiscal Balance -4.0
    - Crisis** Overall Fiscal Balance -6.3
    - Pre-Crisis* Current Account Balance -5.6
    - Crisis** Current Account Balance -5.6
    - Pre-Crisis* Real GDP p.c. growth 2.2
    - Crisis** Real GDP p.c. growth 0.4

  - AM
    - Pre-Crisis* Overall Fiscal Balance -1.9
    - Crisis** Overall Fiscal Balance -3.6
    - Pre-Crisis* Current Account Balance -6.5
    - Crisis** Current Account Balance -2.5
    - Pre-Crisis* Real GDP p.c. growth 2.7
    - Crisis** Real GDP p.c. growth -1.6

  - EM
    - Pre-Crisis* Overall Fiscal Balance -5.0
    - Crisis** Overall Fiscal Balance -9.1
    - Pre-Crisis* Current Account Balance -3.0
    - Crisis** Current Account Balance -4.3
    - Pre-Crisis* Real GDP p.c. growth 2.3
    - Crisis** Real GDP p.c. growth 0.0

  - LIDC
    - Pre-Crisis* Overall Fiscal Balance -3.9
    - Crisis** Overall Fiscal Balance -5.1
    - Pre-Crisis* Current Account Balance -5.8
    - Crisis** Current Account Balance -6.2
    - Pre-Crisis* Real GDP p.c. growth 1.6
    - Crisis** Real GDP p.c. growth 1.1

  - SDS
    - Pre-Crisis* Overall Fiscal Balance -2.6
    - Crisis** Overall Fiscal Balance -3.5
    - Pre-Crisis* Current Account Balance -11.5
    - Crisis** Current Account Balance -9.1
    - Pre-Crisis* Real GDP p.c. growth 2.4
    - Crisis** Real GDP p.c. growth 1.4

  - Note: * t-3 to t-1. ** t to t+2.

- Stylized outcomes and cross-group differences:
  - Economic growth and the fiscal deficit deteriorate substantially with the onset of a crisis; the current account deficit remains large but stable.  
  - AMs suffer the largest growth decline; their pre-crisis years show large external imbalances, which improve afterwards.  
  - EMs experience the largest increases in fiscal deficits around the onset and the largest deterioration in current account balance.  
  - LIDCs have the smallest decline in economic growth, albeit starting from the lowest pre-crisis level.

### Dynamics around crises — event-study and panel findings
- Method: fixed-effects panel regressions with an 11-year event window (j = -5 to +5), following Gourinchas and Obstfeld (2012).  
- Main empirical findings:
  - A fiscal crisis tends to be preceded by loose fiscal policy: robust (above normal times) real expenditure growth in the run-up.  
  - Once the crisis begins, governments aggressively contain expenditure growth — fiscal policy is procyclical as economic conditions weaken.  
  - Budget revenues grow at a weaker pace and the fiscal balance continues to worsen during crisis years.  
  - Expenditure policy is more procyclical in AMs and EMs and less so in LIDCs around crises.  
  - Primary balance deterioration in crisis years is strongest among EMs despite a significant reduction in expenditure growth.  
  - AMs and SDSs face a more temporary deterioration in the primary balance and undo it faster due to a large upfront deceleration in expenditure growth.  
  - Crises signaled by credit events and implicit defaults have the largest increases in the post-crisis budget deficit.  
  - Economic growth is generally higher than in normal times in the run-up, then declines sharply at crisis onset; growth tends to slow in the year preceding the crisis and fall significantly with the onset.  
  - Fiscal crises triggered by implicit default come with more severe economic deterioration than all others.  
  - The growth pattern for credit events is mainly driven by cases involving private creditors; when only public creditors are involved, the decline in growth from pre-crisis years is much smoother and growth converges to normal times.  
  - AMs experience the largest fall in real growth in the first two years of the crisis; EMs have the second largest fall. The fall in growth for these two groups varies between 3 and 6 percentage points in the first two years of the crisis.  
  - Almost half of AMs and EMs experience negative growth in the first and second year of the crisis (Appendix Table 5). About a third of LIDCs face negative growth in the first year of the crisis.  
  - Public debt levels remain higher than normal throughout the crisis period; at the crisis onset public debt ratios rise substantially, especially in AMs and EMs, and fall only gradually afterward.  
  - In LIDCs, debt falls with the start of the crisis (from above-average levels), often reflecting debt defaults.  
  - Crises associated with credit events where official creditors dominate show the most pronounced fall in public debt. Events involving private creditors tend to keep public debt elevated; when only public creditors are affected, debt levels usually fall quickly.  
  - Crises identified by IMF program access or loss of market confidence are usually accompanied by larger external imbalances. IMF programs involve substantial upfront cuts to public expenditure growth and a large improvement in the external current account; post-crisis public debt ratios stabilize around the average for normal periods.  
  - Twin deficits worsen in the crisis run-up, a pattern dominating in AMs and EMs, with the post-crisis current account adjustment being largest in AMs.  
  - Except for SDSs, countries tend to experience a sharp exchange rate depreciation peaking in the year after the start of the crisis.

*Source: Authors' calculations.*

### Box 1. Was the Recent Wave of Fiscal Crises Different?

### Box 1. Was the Recent Wave of Fiscal Crises Different?

### Overview: incidence and timing
- Worldwide, 80 fiscal crises occurred after the onset of the global financial crisis (GFC) in 2008.
- All country groups were affected to a high degree (except SDSs).
- Many of the recent fiscal crises occurred between 2008 and 2014, with a concentration for AMs between 2008 and 2012.

### Advanced Markets (AMs): features and impacts
- Preceded by larger twin deficits relative to previous crises.
- A fifth of AMs went into a fiscal crisis—all of which occurred between 2008 and 2012.
- Crisis duration: on average somewhat shorter than previous crises, but stronger in impact.
- Public debt dynamics:
  - Public debt levels rise, peaking three years after the onset of the fiscal crisis.
  - Peak was more than 50 percentage points of GDP higher than previous fiscal crises.
- Expenditure and external balances:
  - Crisis triggered when expenditure growth was stronger than past experience.
  - Recent crises were preceded by a much larger deterioration in the external current account.
- Growth, inflation, and exchange rates:
  - Economic growth deteriorated significantly only after a few years—by more than 4 percentage points less at its lowest point.
  - Inflation dropped well below previous crises’ levels at the crisis onset.
  - Exchange rates experienced a short-lived sizable stronger depreciation shortly after crisis onset, partly fueled by quantitative easing (QE).

### Emerging Markets (EMs): features and impacts
- Two-fifths of EMs went into a fiscal crisis between 2008 and 2014.
- Affected EMs were mostly financially interconnected countries in Emerging Europe and resource-rich countries in Latin America and the Middle East.
- Pre-crisis position:
  - On average enjoyed a period of strong growth, with lower public debt levels and smaller fiscal deficits than previous episodes.
- Debt dynamics:
  - At crisis onset, debt levels had converged to average levels in other crisis periods and outpaced them a few years after onset (but to a much lesser degree than AMs).
- Interest rates and exchange rates:
  - Partly reflecting AMs’ QE, there was a short-lived but large dip in both effective interest rates and exchange rate depreciation in the first year after crisis onset.
- Growth:
  - Economic growth tended to remain somewhat above levels of previous crises.

### Low-Income Developing Countries (LIDCs): features and impacts
- Two-thirds of LIDCs went into fiscal crisis between 2008 and 2014.
- Starting position:
  - Generally started from a stronger position—especially much lower debt levels (reflecting in part previous debt relief) and higher growth.
- Fiscal and growth dynamics:
  - Despite some acceleration in expenditure growth when the last crisis hit, LIDCs were able to contain fiscal deterioration, partly reflecting stronger real GDP growth relative to other crises.
  - Consequently, debt levels remained well below those of previous crisis episodes.
- Note: LIDCs often show large pre-crisis public debt buildup in some twin crisis episodes, reflecting political instability and economic turbulence.

### Methods underpinning the comparisons
- Findings based on event studies using a fixed-effect difference-in-difference regression; time dummies are interacted with a “this time is different” dummy (value = 1 if a fiscal crisis happens in 2008 or later).
- Twin-crisis analysis uses a fixed-effect DID with separate banking crisis dummy Bi and currency crisis dummy Ci over an 11-year window around a fiscal crisis.
- Long-term impact analysis uses an impulse response function (IRF) approach à la Cerra and Saxena (2008), estimating a fixed-effect AR(p) model with p=1 and q=4; IRFs derived with one-standard-error bands from 1,000 Monte Carlo simulations.

### Twin fiscal-financial crises: amplification effects
- General result: fiscal crises overlapping with banking or currency crises show amplified macro-fiscal deterioration relative to stand-alone fiscal crises.
- Sovereign indebtedness:
  - Fiscal-banking crises: debt ratios increase by an additional 10 percentage points of GDP (impact felt over time).
  - Fiscal-currency crises: debt rises by around an additional 10 percent of GDP relative to a fiscal crisis (effect more immediate but tends to dissipate after some years).
  - Main drivers of higher debt in fiscal-banking crises are AMs and EMs, reflecting fiscal costs such as bank asset purchases and bank capital injections.
- Growth effects:
  - Fiscal-banking crises: growth is lower by an additional 2½ percentage points on average in the year after crisis onset (amplification especially strong in AMs).
  - Fiscal-currency crises: present an additional growth contraction of about 2 to 3.5 percentage points in the first two years (driven by EMs and some AMs).
- Heterogeneity:
  - LIDCs’ twin crises are often preceded by economic turbulence; patterns differ from AMs and EMs.

### Long-term growth and debt impact
- Public debt:
  - On average, public debt returns to similar levels as the pre-crisis trend, but with differences across country groups.
  - AMs: public debt ratio rises by around 10 percentage points of GDP at crisis onset and remains about 4 percentage points of GDP above pre-crisis level in the long term.
  - EMs: experience a rise in public debt but eventually converge to the pre-crisis level.
  - LIDCs and SDSs: public debt first falls but eventually returns to levels similar to pre-crisis periods.
- Real GDP per capita:
  - Fiscal crises are associated with a permanent loss of real GDP of around 2 percent on average.
  - Heterogeneity:
    - AMs and EMs: experience greater output losses during a fiscal crisis but tend to recover half of the initial output losses.
    - LIDCs and SDSs: face a more gradual fall in GDP with no recovery, converging to pre-crisis growth rates over time.
- Case illustrations: Indonesia (1997) shows persistent deviation from pre-crisis trend; Turkey (1998) shows partial recovery toward pre-crisis growth over time.

### Key policy implications and conclusions
- The evidence supports adopting prudent macro-fiscal policies to avoid disruptive fiscal adjustment when a crisis starts.
- Defaulting on debt or financing deficits by printing money should not be seen as costless alternatives to sound policy.
- Crises are generally preceded by loose fiscal policy and twin deficits; once a crisis arrives, countries are likely to tighten fiscal policy in difficult conditions.
- The new comprehensive database extends crisis identification beyond sovereign defaults and broadens country coverage, enabling analysis across all development levels and highlighting LIDCs’ vulnerability.
- Suggested future research directions include better identifying economic and policy precursors to crises, developing early warning indicators, and improving policy options for prevention and management.

*Source: Box 1, wp1786 - Was the Recent Wave of Fiscal Crises Different?*

### Annex 1. Data Sources

### wp1786 - Annex 1. Data Sources

### Coverage and sample
- The database covers 188 countries (all IMF members as of end-2015) and 46 years (1970-2015).
- Country grouping for analysis:
  - Advanced economies (AMs): 35
  - Non-small emerging economies (EMs): 70
  - Non-small low-income developing economies (LIDCs): 50
  - Small developing states (SDSs): 33

### Primary data sources and variable coverage
- Most variables: WEO and its previous vintages.
- Gross debt time series expanded into the past using FAD’s Historical Public Debt Database (in conjunction with WEO). (See Appendix Tables 1 and 5 in the source.)
- Sovereign default / credit event data: Bank of Canada’s CRAG sovereign defaults database (BoC (2016)).
  - BoC (2016) definition of sovereign default includes:
    - debt service not paid on the due date (or within a specified grace period), or payments not made within the time frame specified under a guarantee;
    - or, absent outright payment default, occurrences where creditors incur material economic losses on the sovereign debt they hold, including:
      - (i) agreements between governments and creditors that reduce interest rates and/or extend maturities on outstanding debt;
      - (ii) government exchange offers to creditors where existing debt is swapped for new debt on less-economic terms;
      - (iii) government purchases of debt at substantial discounts to par;
      - (iv) government redenomination of foreign currency debt into new local currency obligations on less-economic terms;
      - (v) swaps of sovereign debt for equity (usually relating to privatization programs) on less-economic terms;
      - (vi) conversion of central bank notes into new currency of less-than-equivalent face value.
  - Coverage: 136 countries from 1970-2015.
  - Gap-closure approach:
    - First identify countries that never defaulted using rating agency reports (especially Moody's and S&P).
    - Then derive lower-bound estimates of sovereign debt in default from separately available components of BoC’s (2016) data (including restructured amounts from Cruces and Trebesch (2013), rescheduled and relieved amounts from WB IDS and WDI).
- IMF financial program data: IMF database covering all members from 1952 to 2015.
- Other account payables (OAP) data (proxy for arrears): Eurostat and OECD data portal (Eurostat prioritized); available for most OECD countries starting in the early 1990s.
- Yields: IFS.
- EMBI: Reuters Datastream.
- CDS Spreads: Bloomberg.
  - For criterion 4 (market confidence) EMBI spreads are prioritized when available.
- Loss of market access dummies:
  - Guscina, Sheheryar, and Papaioannou (2016): for 57 countries from 1990 onwards.
  - For years prior to 1990 and missing observations: Gelos, Sahay, and Sandleris (2004) database of market access for 140 developing countries from 1980 through 2000.
- Currency and banking crisis indicators: Laeven and Valencia (2012).
  - Banking crisis signaled if at least 3 of the following 6 criteria are met:
    - (i) extensive liquidity support;
    - (ii) bank restructuring gross costs;
    - (iii) significant bank nationalization;
    - (iv) significant guarantees;
    - (v) significant asset purchases;
    - (vi) deposit freezes.
  - Currency crisis signaled if the nominal rate of depreciation is at least 30 percent and it is also 10 percentage points larger than the previous year’s rate of depreciation.

### Sample composition (Appendix Table 1)
- Appendix Table 1 lists the countries in each group (AMs, EMs, LICs, SDSs) as defined and used in the analysis. (Full country lists are provided in Appendix Table 1 of the source.)

### Key dataset counts and summary statistics (from Appendix tables)
- Total observations (sample universe reported in Appendix Table 3): 8648.
- Crisis years (overall reported under Criterion 1 in Appendix Table 3): 1715.
- Total crisis episodes reported in Appendix Table 5: 436.
- Percent of episodes with negative real GDP p.c. growth (Appendix Table 5, "Percent of country group episodes with negative real GDP per capita growth"):
  - Total: 24.8 (t-3), 25.5 (t-2), 28.2 (t-1), 40.6 (t), 33.9 (t+1), 29.8 (t+2)
  - AM (25 episodes): 12.0 (t-3), 28.0 (t-2), 28.0 (t-1), 48.0 (t), 60.0 (t+1), 48.0 (t+2)
  - EM (154 episodes): 22.1 (t-3), 20.8 (t-2), 28.6 (t-1), 49.4 (t), 40.3 (t+1), 29.2 (t+2)
  - LIDC (171 episodes): 29.8 (t-3), 28.1 (t-2), 28.1 (t-1), 33.9 (t), 29.2 (t+1), 29.8 (t+2)
  - SDS (86 episodes): 23.3 (t-3), 27.9 (t-2), 27.9 (t-1), 36.0 (t), 24.4 (t+1), 25.6 (t+2)

### Event-study and empirical approaches (figures and methodology notes)
- Event study window: 11-year time window (t-5 to t+5), comparing values during the window to a “normal” period average (following Gourinchas and Obstfeld (2012)).
- One outlier case dropped from event study: Zimbabwe 2000, because the primary balance ratio was reduced by more than 1600 percentage points 4 years after the onset of the fiscal crisis.
- Variables analyzed in event studies and plots include:
  - Public debt ratio (percent of GDP)
  - Effective interest rate (percent)
  - Exchange rate depreciation (percent)
  - Current account balance (percent of GDP)
  - Revenue growth (percent)
  - Expenditure growth (percent)
  - Primary balance (percent of GDP)
  - Real p.c. GDP growth (percent)
  - Inflation rate (percent)
- Event studies disaggregated by country groups (AMs, EMs, LIDCs, SDSs) and by crisis criteria (Credit Events; Official Financing; Implicit Default; Market Confidence).
- Twin-crisis analyses:
  - Twin fiscal-banking crises: banking crisis occurs during [t-1, t+1] where T is the start year of a fiscal crisis; estimates of additional effects on public debt ratio and real GDP p.c. growth obtained via extended event study with difference-in-difference estimations.
  - Twin fiscal-currency crises: currency crisis occurs during [t-1, t+1] where T is the start year of a fiscal crisis; estimates of additional effects on public debt ratio and real GDP p.c. growth similarly derived.
- Monte Carlo simulations for fiscal crisis response:
  - Follow Cerra and Saxena (2008) estimating a fixed-effect AR(p) model to simulate impulse response functions.
  - Figures present the mean of simulations with a one standard deviation corridor for the response of public debt ratio and real GDP p.c. to a one-period exogenous fiscal crisis shock.

*Source: wp1786 - Annex 1. Data Sources (IMF working paper annex content).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2017/wp1786.pdf_