## _wp15176

## Source details

**Canonical URL:** [_wp15176](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15176.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15176.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15176.pdf.json)

---

### Introduction and research focus
- Context: With falling oil and commodity prices, inflation has been declining in advanced economies and is running significantly below targets, raising concerns over the risk of deflation.
- Research question: assess how deflation may buffet already-strained public finances and further complicate fiscal policy by examining the effects of declining prices on fiscal aggregates (debt-to-GDP ratios, government revenue, and expenditure).
- Dataset: an original panel dataset covering a long timeframe (over 150 years) with inflation, growth, and fiscal aggregates—including debt-to-GDP ratios for 21 advanced economies.

### Three central messages (Key findings)
- Finding 1: Deflation negatively affects debt-to-GDP ratios and impacts nominal budgetary variables (government revenue and expenditure). Piloting fiscal policy during deflation can be challenging, similar to managing negative output growth.
- Finding 2: Linkages have evolved over time. Public finance management transitioned from marginal, nominally-driven, cash-constrained regimes to modern governments using large aggregates and indexation mechanisms. Deflation episodes have been relatively rare in the recent period, with the notable exception of Japan in the 1990s.
- Finding 3: Not all deflations are alike — the fiscal impact differs markedly between deflations associated with positive growth and deflations accompanied by recessions.

### Theoretical background — channels through which deflation affects public finances
- Debt dynamics and Fisher effect:
  - Debt dynamics combine the initial debt stock, the real interest rate, the primary balance, and stock-flow adjustments; real interest rate = nominal interest rate − actual inflation rate.
  - For any given debt stock and real growth rates, deflation mechanically increases the debt-to-GDP ratio by lowering nominal GDP.
- Primary balance — revenue channels:
  - Seigniorage: deflation reduces seigniorage for a given level of real money balances; one additional point of inflation would lead today to about 0.12 percent of GDP annually additional seigniorage revenue for G7 countries (Akitoby and others, 2014).
  - Tax system features: under a fully proportional tax system, deflation would have no impact on the revenue-to-GDP ratio; progressivity and non-indexed tax brackets cause deflation to curb revenue ratios; price-inelastic components (excises, non-tax revenues) are inertial and can boost revenue as percent of GDP during deflation.
  - Behavioral and composition effects: consumption shifts and relative price moves can alter revenue responsiveness.
- Primary balance — expenditure channels:
  - Nominal rigidities and political economy: downward rigidities in wages, social transfers, and contractual commitments make cutting nominal spending difficult; freezing nominal spending often raises expenditure-to-GDP ratios.
  - Contractual and multi-year provisions delay transmission of deflation to some capital and recurrent spending.
- Dependence on debt characteristics:
  - Impact depends on maturity structure, currency denomination, and share of price-indexed bonds.

### Empirical strategy — methodology and identification challenges
- Approach:
  - Estimate autoregressive models in first differences for fiscal ratios (changes in debt, primary deficit, primary expenditure, revenues) expressed as percent of GDP.
  - Models include current and lagged inflation and output growth, country and year fixed effects, and dummies for structural breaks: pre-Great Depression (1851–1928), Great Depression (1929–34), post-WWII (1946–2013).
  - Inflation decomposed into positive and negative parts; zero used as natural threshold for deflation (alternative threshold of 1 percent tested with similar results).
- Lags and estimation:
  - Optimal lag length tested with Akaike, Hannan-Quinn, and Schwarz criteria; optimal lag length found between 1 and 3.
  - All estimates conducted with 3 lags whenever possible; autocorrelation and heteroskedasticity corrected.
- Identification challenges and mitigation:
  - Reverse causality, omitted variables (war, colonization, political instability), and multicollinearity acknowledged; country and year fixed effects and structural-break dummies used to mitigate bias.
  - Instrumental approaches were explored but suitable instruments over the long historical span were not found; GMM was ruled out due to bias when time dimension exceeds cross-section.

### Empirical strategy — dataset
- Coverage and sources:
  - Inflation: CPI-based inflation from Bordo and Filardo (2005) through 1997, complemented by IMF WEO; Warren Weber’s dataset used where Bordo and Filardo missing (prices and output 1810–1995).
  - Growth: real GDP growth from Bordo and Filardo (2005), complemented by WEO.
  - Debt ratios: Abbas and others (2010), supplemented with WEO.
  - Fiscal flows and interest payments: Mauro and others (2013).
  - Panel: unbalanced panel of data for 21 developed economies dating from as early as 1851. Countries include Australia, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Italy, Japan, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the United States.
- Definitions and treatment:
  - Deflation defined as negative inflation; low inflation defined as inflation between 0 and 2 percent; recession defined as negative annual growth rate in real GDP.
  - Exclude extreme outliers in price dynamics (inflation above 100 percent).

### Descriptive statistics (preserved exactly)
- Inflation: Mean 4.08, Std. Dev. 8.1, Minimum -20.28, Maximum 796.95, N 42380
- Growth: Mean 3.04, Std. Dev. 4.1, Minimum -21.70, Maximum 93.10, N 42380
- Debt-to-GDP: Mean 52.93, Std. Dev. 8.2, Minimum 0.66, Maximum 5269.79, N 42380
- Primary balance-to-GDP: Mean 0.5, Std. Dev. 4.5, Minimum -47.67, Maximum 20.24, N 52380
- Revenue-to-GDP: Mean 23.7, Std. Dev. 16.3, Minimum 0.77, Maximum 165.27, N 32380
- Primary expenditure-to-GDP: Mean 23.1, Std. Dev. 16.2, Minimum 0.56, Maximum 66.42, N 82380
- Δ Debt-to-GDP: Mean 0.4, Std. Dev. 6.1, Minimum -52.52, Maximum 941.30, N 32380
- Δ Primary balance-to-GDP: Mean 0.0, Std. Dev. 2.5, Minimum -26.38, Maximum 29.28, N 02380
- Δ Revenue-to-GDP: Mean 0.3, Std. Dev. 2.0, Minimum -12.06, Maximum 227.47, N 72380
- Δ Primary expenditure-to-GDP: Mean 0.3, Std. Dev. 2.9, Minimum -21.57, Maximum 25.86, N 02380

### Historical evidence — main empirical findings
- Overall:
  - Historical evidence suggests deflation increases the debt-to-GDP ratio primarily through a worsening of interest rate-growth differentials, while primary balances remain broadly unaffected on average.
- Inflation effects on debt-to-GDP:
  - A 1 percentage point increase in the inflation rate (when inflation is positive) translates into a reduction of the debt ratio of 0.15 percent of GDP in a year.
  - When inflation is negative (deflation), it leads to an increase in the public debt ratio of 0.19 percentage point in a year.
  - There is a delayed positive effect of inflation on the debt ratios after two years.
  - Change in debt-to-GDP ratio is persistent; the dependence coefficient is positive and below unity in absolute terms.
- Growth and interest effects:
  - A 1 percentage point increase in growth leads to a decline of 0.40 percentage point in the debt-to-GDP ratio.
  - Impact of growth in the following year is positive, estimated at 0.14 percentage point.
  - A 100 basis points increase in the nominal rate raises the debt-to-GDP ratio by 0.65 percentage point; delayed impact of nominal rate changes is negative with lower magnitude.
- Deflation heterogeneity:
  - State of deflation permanently increases debt ratios: debt-to-GDP ratio increases by almost 1.7 percent a year.
  - Recessionary deflation (deflation combined with negative growth): debt-to-GDP ratio increases by 3.2 percent a year.
  - Expansionary deflation (deflation combined with positive growth): no impact on debt-to-GDP ratio.
  - Public debt-to-GDP ratios grow on average by 1.7 percentage points of GDP a year during deflation periods, three times faster than the sample average.

### Primary balances, revenues, and expenditures — empirical patterns
- Aggregate result:
  - Higher inflation reduces both revenue and expenditure ratios by similar magnitudes—0.04 percentage point of GDP—leaving the primary balance broadly unchanged.
  - Deflation has no significant impact on the primary balance and its components as ratios of GDP in many specifications; deterioration occurs when deflation coincides with recession.
- Nominal (pseudo-nominal) changes:
  - Both nominal expenditure and nominal revenue are positively correlated with inflation.
  - Positive inflation boosts nominal revenue and expenditure; deflation does not deflate them as much — indicative of downward rigidities in expenditures and in tax provisions.
  - Negative effect of deflation on nominal revenue is larger than on nominal expenditure; discrepancy increases over time.
  - Regime-dummy results:
    - Expansionary deflation: primary balance deteriorates via declining revenue and rising expenditure (expenditure coefficient may lose significance).
    - Recessionary deflation: expenditure cuts exceed revenue shortfalls (reflecting Great Depression episodes).
    - Restricting sample to 1946–2013: only the negative response of revenue persists, suggesting modern reluctance to downsize spending when revenues fall.

### Simulation exercise — stylized Euro area model (setup and assumptions)
- Purpose: incorporate compositional fiscal effects, indexation mechanisms, fiscal rules, and modern policy practices under “unchanged policy” assumption; explore asymmetric impact of deflation on revenue and expenditure and cost of unanticipated shocks.
- Stylized component shares (percent of GDP, Table 9):
  - Income tax revenues: 14
  - Social contributions: 11
  - Goods and services tax: 10
  - Excises and other tax revenues: 6
  - Non-tax revenues: 5
  - Wage bill: 11
  - Intermediate consumption: 6
  - Social expenditures: 25
  - Investment and capital transfers: 3
  - Other primary expenditure: 3
- Shock simulated: 2 percentage point inflation shock in 2015 and 2016 relative to baseline (baseline: April 2015 WEO forecasts). Baseline inflation over 2015–16 averages 0.5 percent a year.
- Key behavioral assumptions:
  - Asymmetric expenditure response: more elastic upward than downward due to downward rigidities.
  - Revenue elasticities: income tax revenues adjust with a one-year lag for part of the base; social contributions elasticity around one; taxes on goods and services vary proportionally with nominal GDP; excises and some revenues assumed broadly immune.

### Simulation results — stylized quantitative outcomes (preserved exactly)
- Shock types: persistent 2 percentage point downward surprise; persistent 2 percentage point upward surprise; distinction between surprises and anticipated shocks.
- Primary expenditure responses (deviations from baseline by 2019):
  - Unexpected downward 2 percentage point inflation shock → increase in the primary expenditure-to-GDP ratio of 1.4 percentage points by 2019.
  - Unexpected upward 2 percentage point inflation shock → primary expenditure ratio 0.3 percentage point lower by 2019.
  - Expected (anticipated) downward shock → smaller effect on expenditure: 1.2 percent of GDP.
- Revenue responses (deviations from baseline by 2019):
  - Permanent 2 percentage point downward inflation shock → revenue ratio reaches 0.4 percent of GDP by 2019.
  - Permanent 2 percentage point upward inflation shock → revenue ratio around -0.4 percent of GDP by 2019.
  - For an upward shock, revenue ratio decreases more when the shock has been correctly anticipated than when it comes as a surprise because of bracket indexation; indexation assumed not to take place downward.
- Primary balance and debt outcomes:
  - A 2 percentage point decline in inflation in 2015–16 implies a deterioration in the primary balance of 0.9 percentage point in 2019 (0.8 if perfectly anticipated).
  - Disinflationary shock → debt ratios permanently increase by 6–7 percentage points of GDP relative to baseline over a five year period.
  - Inflationary shock → debt ratios temporarily decrease by 0–3 percentage points of GDP, depending on the degree of anticipation.
- Mechanisms amplifying debt effects:
  - Stock effect: higher initial debt increases sensitivity to price variations.
  - Accumulation effect: repeated shocks build up through debt accumulation.
  - Interest bill effect: small and persistent, opposite sign to the effect on the primary balance.

### Lessons for modern policymakers and policy recommendations
- Recognize asymmetrical fiscal behavior in deflationary episodes and consider deliberate countercyclical support when appropriate, noting that such support increases fiscal deficits and should not be passive.
- Act rapidly in budget execution to address unexpected downward revisions of inflation to avoid unwanted ratchet effects on government expenditure.
- Revisit sophisticated fiscal policy tools and rules given low inflation/deflation risk, including examining nominally expressed multiyear expenditure ceilings for downward revision when inflation is below expectations.
- Consider increasing the share of inflation-indexed bonds in the government’s financing mix to alleviate negative effects of deflation on debt ratios.
- Tailor approaches to country specifics, including political economy dimensions; distinguish whether deflation is supply-driven or demand-driven when designing responses.

### Appendix A — Japan (1990–2013) highlights
- Primary expenditure and revenue movements:
  - Primary expenditure-to-GDP ratio rose steadily by 12 percent during 1990–2013.
  - Revenue-to-GDP ratio declined by 4 percent during 1990–2002 and returned to its initial level by 2013.
- Debt dynamics:
  - Public debt rose from 69 percent of GDP in 1990 to 243 percent in 2013.
  - The sole denominator effect of inflation explains roughly one fifth of the debt increase between 1990 and 2013: 32 percentage points out of 174 percentage points.
  - Nominal interest rates fell during the deflationary period but remained above the nominal growth rate on average.
- Institutional note:
  - Downward indexation was applied late and only for a few years; pension indexation that would have called for absolute declines was not applied from 2000–02 and was suspended after 2004.

### Limitations, caveats, and conditional factors
- Historical data reflect ex-post outturns that incorporate discretionary measures adopted in response to deflation; structural changes in policymaking and the relative rarity of recent deflation episodes limit direct inference for modern contexts.
- Dataset contains few modern episodes of deflation; pre–World War I institutional settings differed (limited scope for intervention, tax regimes not capturing nominal GDP changes effectively).
- Harm from deflation is greater when accompanied by economic recession or stagnation; distinguishing supply versus demand origins of deflation is crucial but difficult.
- Expectations and the distinction between surprise and anticipated shocks materially affect fiscal outcomes.

*Excerpt from the supplied IMF working paper content.*

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

### References

### Introduction and research focus
- With falling oil and commodity prices, inflation has been declining in advanced economies and is running significantly below targets, raising concerns over the risk of deflation.
- Research question: assess how deflation may buffet already-strained public finances and further complicate fiscal policy by examining the effects of declining prices on fiscal aggregates (debt-to-GDP ratios, government revenue, and expenditure).
- Dataset: an original panel dataset covering a long timeframe (over 150 years) with inflation, growth, and fiscal aggregates—including debt-to-GDP ratios for 21 advanced economies.

### Theoretical context and literature
- Deflation mechanically increases debt-to-GDP ratios because debt is a stock and downward rigidities in sovereign interest rates compound the increase.
- Nominal flow variables (revenue, nominal GDP, and to a smaller extent expenditure) tend to decrease mechanically in deflation.
- Prior literature focus areas:
  - Fiscal stances’ effect on price dynamics (Catao and Terrones 2005).
  - Fiscal consequences of very high inflation (Oliveira, 1967; Tanzi, 1977; Aghevli and Kahn, 1978; Heller, 1980).
  - Role of fiscal policy in exiting deflation (Auerbach and Obstfeld, 2004; Cochrane, 2011).
  - Link between deflation and recession, with categorizations of deflation: good (supply-driven), bad (demand-collapse, e.g., Japan in the 1990s), and ugly (severe recessions, e.g., early 1930s Great Depression) (Borio and Filardo 2004, 2005).

### Empirical approach and scope
- Empirical strategy: historical analysis using the long-run panel to examine whether the impact of deflation on fiscal aggregates is asymmetrical relative to inflation.
- The paper focuses on consumer price inflation; asset price deflation is noted as likely less important for fiscal accounts than GDP and consumption prices.
- The paper adopts a positive approach and does not analyze the optimal fiscal response to deflation.

### Key findings (three central messages)
- Finding 1: Deflation negatively affects debt-to-GDP ratios and impacts nominal budgetary variables (government revenue and expenditure). Piloting fiscal policy during deflation can be challenging, similar to managing negative output growth.
- Finding 2: Linkages have evolved over time. Public finance management transitioned from marginal, nominally-driven, cash-constrained regimes to modern governments using large aggregates and indexation mechanisms. Deflation episodes have been relatively rare in the recent period, with the notable exception of Japan in the 1990s.
- Finding 3: Not all deflations are alike — the fiscal impact differs markedly between deflations associated with positive growth and deflations accompanied by recessions.

### Japan case and quantitative highlights
- Japan’s deflation was associated with deteriorating fiscal aggregates, mostly the debt-to-GDP ratio.
- Public debt doubled, due primarily to a snowball effect (an unfavorable differential between interest rate and growth).
- The denominator effect of deflation explains roughly one fifth of the debt increase.
- The impact of deflation on the primary deficit was blurred by demographic changes and policy responses (see Appendix A).
- Expenditure-to-GDP ratios increased largely because of rising age-related spending and explicit downward rigidities.
- The impact on the tax-to-GDP ratio is difficult to identify because new tax measures were introduced simultaneously with a shift in the tax base; protracted and anticipated deflation allowed partial offsetting in annual budgets.

### Caveats and limitations
- Historical data reflect ex-post outturns that incorporate discretionary measures adopted in response to deflation, making underlying trends hard to identify.
- Historical records do not fully account for structural changes in policymaking introduced in recent decades, given the relative rarity of recent deflation episodes.
- A prospective simulation exercise (under a no-policy change assumption and including expectations and modern government features) confirms that piloting fiscal policy amid deflation can be complicated; composition effects within the fiscal balance and the role of expectations deserve particular attention.

### Paper organization (structure)
- Section II: theoretical discussion of how deflation could affect fiscal aggregates.
- Section III: description of empirical methodology and the historical dataset assembled to capture sufficient deflation episodes.
- Section IV: presentation of empirical results.
- Appendices include: A. Deflation in Japan; B. Unit Root Tests; C. Optimal Lag Tests.

* _wp15176 - References_

### Section V attempts to put in a narrative, historical perspective. Section V closes by

### _wp15176 - Section V attempts to put in a narrative, historical perspective. Section V closes by

### Theoretical background — channels through which deflation affects public finances
- General:
  - Public finances are vulnerable to deflation through effects on debt ratios (Fisher, 1933) and via impacts on primary balances through revenue and expenditure channels.
  - Debt dynamics combine the initial debt stock, the real interest rate, the primary balance, and stock-flow adjustments. Real interest rate is defined as the difference between the nominal interest rate and the actual inflation rate.
- Primary balance — revenue:
  - Seigniorage:
    - Under a fiat money system and without monetary policy action, seigniorage revenue equals the product of the inflation rate and real money balances (inflation tax).
    - Deflation reduces seigniorage for a given level of real money balances, creating a “deflation subsidy”.
    - If deflation increases holdings of real money balances, the effective seigniorage tax base may augment and seigniorage revenue could increase.
    - Recent evidence suggests potential gains of seigniorage are limited in today’s advanced economies because base money is small relatively to GDP.
    - Simulation evidence cited: one additional point of inflation would lead today to about 0.12 percent of GDP annually additional seigniorage revenue for G7 countries (Akitoby and others, 2014).
  - Tax system features:
    - Under a fully proportional tax system, deflation would have no impact on the revenue-to-GDP ratio because nominal revenue and GDP would move proportionally.
    - Progressivity and non-indexed tax brackets: deflation tends to curb revenue ratios by moving taxpayers to lower tax brackets if tax brackets are not perfectly indexed.
    - Widespread nominal tax exemptions increase in relative cost when prices fall, reducing revenue-to-GDP ratios.
    - Price-inelastic components (excises, non-tax revenues) are more inertial and can boost revenue as a percent of GDP during deflation.
    - Behavioral effects: consumption may shift toward higher-scale goods during deflation, which tend to be more heavily taxed, potentially improving revenue.
    - If consumption prices drop more rapidly than the GDP deflator, the revenue ratio will tend to rise.
    - Central bank quantitative easing against protracted deflation can generate seigniorage revenue.
    - Full indexation of tax brackets and thresholds would mechanically preserve tax pressure during deflation and make revenue-to-GDP ratios immune to deflation.
- Primary balance — expenditure:
  - Nominal rigidities and political economy:
    - Public expenditure is generally more sensitive to deflation than revenue because of nominal rigidities (difficulty reducing wages and social transfers when prices are falling).
    - Freezing nominal spending is often the politically feasible option, leading to increases in expenditure-to-GDP ratios; especially pronounced for social transfers (pensions and other benefits) and wages.
    - Indexation mechanisms for pensions could call for absolute declines to preserve sustainability but may be suspended (example: Japan indexation suspended during deflation period).
  - Contractual and multi-year provisions:
    - Multiyear agreements and lagged price indexation delay transmission of deflation to capital spending and some recurrent expenses specified in multi-annual contracts (maintenance, IT outsourcing).
  - Budget design:
    - Budgets prepared and executed in nominal terms can delay within-year adjustments to unexpected deflation shocks.
- Debt:
  - Mechanical and behavioral channels:
    - For any given debt stock and real growth rates, deflation mechanically increases the debt-to-GDP ratio by lowering nominal GDP.
    - Primary balances can deteriorate unexpectedly in deflationary environments, further increasing the debt burden.
    - For any given nominal interest and real growth rates, deflation raises the real value of the interest bill if nominal interest rates are sticky or deflation is not anticipated.
  - Dependence on debt characteristics:
    - The impact depends on maturity structure, currency denomination, and the share of price-indexed bonds in total debt.
  - Debt dynamic equation:
    - Year-on-year changes in the debt-to-GDP ratio link to the existing debt stock via nominal interest rates, inflation, output growth, the primary balance, and stock-flow adjustments.
    - A positive value of the term capturing compounded effects implies the debt stock increases over time even with a primary surplus.

### Empirical strategy — methodology and identification challenges
- Methodology:
  - Main approach: estimate autoregressive models in first differences for fiscal ratios (changes in debt, primary deficit, primary expenditure, revenues) expressed as percent of GDP.
  - Equations (2)–(4): changes in fiscal ratios modeled as functions of current and lagged inflation and output growth; include country and year fixed effects; include dummies for structural breaks corresponding to pre-Great Depression (1851–1928), Great Depression (1929–34), and post-WWII (1946–2013).
  - Distinction between inflation regimes: inflation term broken into positive and negative inflations; also use dummy variables to capture different regimes of inflation and growth (zero used as natural threshold for deflation; alternative threshold of 1 percent tested with similar results).
  - Lags and estimation details:
    - Optimal lag length tested using Akaike, Hannan-Quinn, and Schwarz information criteria; optimal lag length found between 1 and 3.
    - All estimates conducted with 3 lags whenever possible to capture autocorrelation structure.
    - Autocorrelation and heteroskedasticity corrected using standard practices.
- Identification challenges and mitigation:
  - Reverse causality: fiscal policy can affect inflation; instrumental variable approaches explored but good instruments over long historical span are difficult to find; lagged effects of inflation included to mitigate simultaneity bias.
  - Instruments tried unsuccessfully included lagged inflation and an index of imported inflation; effective exchange rates unavailable for the full period; institutional variables (e.g., central bank independence, colonial status) difficult to construct historically.
  - Omitted variable bias: war, colonization, political instability may correlate with fiscal performance and inflation; country and year fixed effects and structural-break dummies used to mitigate bias.
  - Multicollinearity: potential between inflation and other controls acknowledged; framework kept deliberately simple and abstracts from second-order cross-linkages among inflation, interest rates, and growth.
  - GMM ruled out due to bias when time dimension exceeds cross-sectional dimension in historical data.

### Empirical strategy — dataset
- Data sources and coverage:
  - Inflation: CPI-based inflation from Bordo and Filardo (2005) through 1997, complemented by IMF WEO; Warren Weber’s dataset used where Bordo and Filardo missing (prices and output 1810–1995).
  - Growth: real GDP growth rates from Bordo and Filardo (2005), complemented by WEO.
  - Debt ratios: Abbas and others (2010), supplemented with WEO.
  - Fiscal flows (primary balances and components, interest payments): Mauro and others (2013).
  - Panel: unbalanced panel of data for 21 developed economies dating from as early as 1851. Countries include Australia, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Italy, Japan, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the United States.
- Definitions and data treatment:
  - Deflation defined as negative inflation; low inflation defined as inflation between 0 and 2 percent; recession defined as negative annual growth rate in real GDP.
  - Exclude extreme outliers in price dynamics, such as hyperinflation episodes (inflation above 100 percent).
- Descriptive statistics (as reported):
  - Average inflation rate: 4 percent; standard deviation: 8.1 percent.
  - Average growth rate: 3 percent; standard deviation: 4.1 percent.
  - Average public debt-to-GDP ratio: 53 percent.
  - Average primary balance: surplus of 0.5 percent of GDP.
  - Average revenue: around 23 percent of GDP.
  - Average primary expenditure: around 23 percent of GDP.
  - Share of sample observations in deflation: about one fifth overall; this ratio shrinks significantly in the most recent period.
  - Debt dynamics during deflation: public debt-to-GDP ratios grow on average by 1.7 percentage points of GDP a year during deflation periods, three times faster than the sample average.
  - Primary balance tendency: on average tends to worsen moderately during deflation, mostly when combined with recession, driven by stalling revenue-to-GDP and increasing expenditure-to-GDP ratios.

### Evidence from the historical records — main findings
- Overall result:
  - Historical evidence suggests deflation increases the debt-to-GDP ratio primarily through a worsening of interest rate-growth differentials, while primary balances remain broadly unaffected on average.
- Debt:
  - Empirical estimates of equation (2) identify a strong and immediate impact of price dynamics on debt-to-GDP ratios.
  - Estimated coefficients are significant at the 5 percent level and align with expectations in the debt accumulation framework: higher inflation helps curb debt-to-GDP ratios.
  - (The source text reports the empirical finding and significance; the numerical magnitude of the contemporaneous effect is not fully reproduced in the supplied excerpt.)
- Primary balances and fiscal components:
  - Primary balances are found to remain broadly unaffected in many specifications, though deterioration occurs when deflation coincides with recession.
  - The deterioration in fiscal balances during deflation is driven mainly by stalling revenues-to-GDP and rising expenditure-to-GDP ratios, reflecting nominal rigidities and non-indexation effects.

*Italic source attribution: Excerpt from the supplied IMF working paper content.*

### 0.15 percentage point. There is a delayed positive effect of inflation on the debt ratios after

### _wp15176 - 0.15 percentage point. There is a delayed positive effect of inflation on the debt ratios after

### Inflation effects on debt-to-GDP
- A 1 percentage point increase in the inflation rate (when inflation is positive) translates into a reduction of the debt ratio of 0.15 percent of GDP in a year.
- There is a delayed positive effect of inflation on the debt ratios after two years.
- The change in debt-to-GDP ratio is persistent; the dependence coefficient is positive and below unity in absolute terms.

### Deflation versus positive inflation (estimated equation (3), Table 2, column (2))
- When inflation is negative (deflation), it leads to an increase in the public debt ratio of 0.19 percentage point in a year.
- When inflation is positive, a 1 percentage point increase reduces the debt ratio by 0.15 percentage point in a year.
- Stability checks over three historical periods (1851–1928; 1929–34; after World War II) indicate deflation was associated with worsening debt-to-GDP ratios mostly during periods prior to World War II, when deflation episodes were more frequent.

### Other control variables and dynamics
- Higher economic growth reduces debt build-up: a 1 percentage point increase in growth leads to a decline of 0.40 percentage point in the debt-to-GDP ratio.
- The impact of growth in the following year is positive, estimated at 0.14 percentage point.
- A higher nominal interest rate is associated with faster debt build-up: a 100 basis points increase in the nominal rate raises the debt-to-GDP ratio by 0.65 percentage point.
- The delayed impact of nominal interest rate changes is negative, but with a lower magnitude.

### Descriptive statistics (Table 1)
- Inflation: Mean 4.08, Std. Dev. 8.1, Minimum -20.28, Maximum 796.95, N 42380
- Growth: Mean 3.04, Std. Dev. 4.1, Minimum -21.70, Maximum 93.10, N 42380
- Debt-to-GDP: Mean 52.93, Std. Dev. 8.2, Minimum 0.66, Maximum 5269.79, N 42380
- Primary balance-to-GDP: Mean 0.5, Std. Dev. 4.5, Minimum -47.67, Maximum 20.24, N 52380
- Revenue-to-GDP: Mean 23.7, Std. Dev. 16.3, Minimum 0.77, Maximum 165.27, N 32380
- Primary expenditure-to-GDP: Mean 23.1, Std. Dev. 16.2, Minimum 0.56, Maximum 66.42, N 82380
- Δ Debt-to-GDP: Mean 0.4, Std. Dev. 6.1, Minimum -52.52, Maximum 941.30, N 32380
- Δ Primary balance-to-GDP: Mean 0.0, Std. Dev. 2.5, Minimum -26.38, Maximum 29.28, N 02380
- Δ Revenue-to-GDP: Mean 0.3, Std. Dev. 2.0, Minimum -12.06, Maximum 227.47, N 72380
- Δ Primary expenditure-to-GDP: Mean 0.3, Std. Dev. 2.9, Minimum -21.57, Maximum 25.86, N 02380

### Figures and regime comparison highlights
- Figure 1 contrasts changes in fiscal aggregates under Positive inflation and Deflation regimes across Overall, Expansionary, and Recessionary contexts for:
  - 1. Debt (scale shown from -1.00 to 7.00 with plotted values indicating negative and positive impacts by regime)
  - 2. Primary Balance (scale shown from -0.60 to 0.20 with plotted values indicating regime differences)
  - 3. Primary Expenditure (scale shown from 0.00 to 0.40 with plotted values indicating regime differences)
- Source for figures: Authors

*Source: _wp15176 - 0.15 percentage point. There is a delayed positive effect of inflation on the debt ratios after (PDF).*

### 3. Revenue

### 3. Revenue

### Deflation and Debt
- State of deflation permanently increases debt ratios: debt-to-GDP ratio increases by almost 1.7 percent a year.
- Effect heterogeneity by growth regime:
  - Recessionary deflation (deflation combined with negative growth): debt-to-GDP ratio increases by 3.2 percent a year.
  - Expansionary deflation (deflation combined with positive growth): no impact on debt-to-GDP ratio.
- Historical context:
  - Before the Great Depression, debt-to-GDP ratios rose by almost 0.3 percentage point a year; during the depression, they rose by 0.5 percentage point.
- Literature evidence summarized:
  - Effects of deflation depend on growth regime; some historical deflations were “primarily good” for the real economy, while “bad” or “ugly” deflations (those combined with recessions) have adverse fiscal effects.

### Primary Balance
- Aggregate result: higher inflation reduces both revenue and expenditure ratios by similar magnitudes—0.04 percentage point of GDP—leaving the primary balance broadly unchanged.
- Direct effect of deflation on primary balance:
  - Deflation has no significant impact on the primary balance and its components as ratios of GDP (Table 4 findings).
  - The contraction in revenue and primary expenditure ratios is attributed to positive inflation; revenue and primary expenditure ratios appear insensitive to deflation.
- Low inflation (0–2 percent) versus deflation:
  - Low inflation has no specific effect on debt.
  - Deflation increases the debt-to-GDP ratio, while inflation above 2 percent curbs it.
  - Low inflation (between 0 and 2 percent) has no significant effect on the primary balance.

### Beyond Ratios: Nominal Changes in Fiscal Aggregates
- Rationale: fiscal ratios can mute the numerator effect (nominal fiscal variables) via the denominator (GDP) effect; authors build “pseudo-nominal” changes (equation (5)) to proxy nominal increases measured in terms of previous year GDP.
- Three robust findings (also hold when deflation defined as inflation below 1 percent):
  1. Both nominal expenditure and nominal revenue are positively correlated with inflation.
     - Revenue and expenditure are historically more responsive to inflation than to deflation (Tables 6–7, columns (3)).
     - Positive inflation boosts nominal revenue and expenditure; deflation does not deflate them as much — indicative of downward rigidities in expenditures and in tax provisions.
  2. Negative effect of deflation on nominal revenue is larger than on nominal expenditure, and this discrepancy increases over time.
     - Coefficients for expenditure consistently lower than those for revenue across historical slices.
     - Result: governments should experience downward pressures on fiscal deficits during deflation, but significant effects on primary balances are not detected.
  3. Regime-dummy results:
     - Deflation associated with lower nominal revenues and lower nominal expenditure (Table 8).
     - Expansionary deflation: primary balance deteriorates via declining revenue and rising expenditure (expenditure coefficient may lose significance).
     - Recessionary deflation: expenditure cuts exceed revenue shortfalls (result may reflect Great Depression episodes).
     - Restricting sample to 1946–2013: only the negative response of revenue persists, suggesting modern reluctance to downsize spending when revenues fall.

### Lessons for Modern Policymakers
- Limitations of historical approach:
  - Impact of deflation on fiscal aggregates is difficult to disentangle from other parameters (wars, evolving fiscal structures, policy changes).
  - Primary balances as ratios of GDP appear unaffected by deflation in historical records, but this masks small nominal impacts that other parameters can silence.
- Structural evolution affecting sensitivity to deflation:
  - Modern governments are larger, more service-oriented, and fiscal frameworks have become more sophisticated (e.g., second generation fiscal rules).
  - Sensitivity of nominal fiscal aggregates to deflation is higher in recent times than earlier periods.
- Revenue-side structural changes:
  - Shift from excises and customs to VATs: example chronology for the United Kingdom—share of customs and excise in revenue was >50 percent in the 1800s, around 25 percent in 1965, and 10 percent nowadays.
  - VATs now account for about 20 percent of government revenue in OECD countries, up from 3 percent in the 1960s; VAT collections are more sensitive to nominal GDP changes.
- Expenditure-side structural changes:
  - Social transfers and public services have expanded since World War I; spending on public services now much larger than core government functions in advanced economies.
  - Governments more reluctant to cut spending (notably social transfers) in deflation—downward nominal rigidities expected to be stronger in modern times.
- Practical implication: modern fiscal frameworks and indexation mechanisms can amplify fiscal sensitivity to price shocks and complicate adjustment in deflationary episodes.

### A Simulation Exercise (Stylized Model for the Euro Area)
- Purpose: incorporate compositional effects of the fiscal balance, indexation mechanisms, fiscal rules, and modern policy practices under “unchanged policy” assumption; explore (i) asymmetric impact of deflation on revenue and expenditure and (ii) cost of unanticipated shocks.
- Key modeling choices:
  - Primary balance decomposed into components with shares of GDP (Eurostat 2013 weights used in Table 9).
  - Nominal impulse responses to inflation shocks specified with contemporaneous and lagged pass-through parameters (equation (6)); allow for anticipated versus surprise shocks.
  - Asymmetric expenditure response: assumed more elastic upward than downward due to downward rigidities (wages, social benefits, intermediate consumption).
  - Revenue assumptions:
    - Income tax revenues (personal and corporate) adjust to inflation with a one-year lag for part of the base.
    - Social contributions: average elasticity of one to inflation, with 10 percent levied on public wage bill.
    - Taxes on goods and services (including VAT) vary proportionally with nominal GDP.
    - Excises, taxes on immovable property, capital and non-tax revenues assumed broadly immune to price shocks.
- Stylized component shares (Table 9, percent of GDP):
  - Income tax revenues: 14
  - Social contributions: 11
  - Goods and services tax: 10
  - Excises and other tax revenues: 6
  - Non-tax revenues: 5
  - Wage bill: 11
  - Intermediate consumption: 6
  - Social expenditures: 25
  - Investment and capital transfers: 3
  - Other primary expenditure: 3
- Scenario setup:
  - Baseline: April 2015 World Economic Outlook forecasts.
  - Shock simulated: 2 percent inflation shock in 2015 and 2016 relative to baseline.
  - Baseline inflation over 2015–16 averages 0.5 percent a year; both upward and downward shocks simulated to assess asymmetry and surprise versus anticipated effects.
- Stylized conclusions (from model description and prior empirical results):
  - Asymmetric nominal responses imply that a downward (deflationary) shock can reduce nominal revenues more than nominal expenditures fall, exerting upward pressure on deficits and debt, particularly if shocks are unanticipated.
  - Anticipation and indexation rules materially alter fiscal outcomes: anticipated upward shocks allow indexation to adjust, whereas surprise downward shocks accentuate rigidities.
  - Modern fiscal structures (higher VAT shares, larger social spending, wage indexation) increase the potential fiscal cost of deflation relative to historical periods.

*Source: IMF staff analysis in "3. Revenue" (extracted from the supplied chapter content).*

### 2.5 percent in one scenario and falls to -1.5 percent in the other one. Moreover, to account for

### _wp15176 - 2.5 percent in one scenario and falls to -1.5 percent in the other one. Moreover, to account for 

### Simulation results: inflation shocks, expenditures, and revenues
- Shock specification and anticipation:
  - A persistent 2 percentage point downward surprise in inflation.
  - A persistent 2 percentage point upward surprise in inflation.
  - Distinction between pure surprises (unexpected shocks) and shocks that governments can anticipate.
- Primary expenditure response (deviations from baseline by 2019):
  - Unexpected downward 2 percentage point inflation shock → increase in the primary expenditure-to-GDP ratio of 1.4 percentage points by 2019.
  - Unexpected upward 2 percentage point inflation shock → primary expenditure ratio 0.3 percentage point lower by 2019 (ex post indexation mechanisms kick in).
  - Expected (anticipated) shock → smaller effect on expenditure: 1.2 percent of GDP (downward shock), because of governments’ reluctance to adjust downward; if inflation shock were on the upside, its impact would almost cancel out by 2019.
- Revenue response (deviations from baseline by 2019):
  - Permanent 2 percentage point downward inflation shock → revenue ratio reaches 0.4 percent of GDP by 2019.
  - Permanent 2 percentage point upward inflation shock → revenue ratio around -0.4 percent of GDP by 2019.
  - For an upward shock, revenue ratio decreases more when the shock has been correctly anticipated than when it comes as a surprise, because of bracket indexation. Since indexation is assumed not to take place downward, the effect of a lower inflation is the same whether expected or not.

### Fiscal balance and debt outcomes
- Primary balance:
  - A 2 percentage point decline in inflation in 2015–16 implies a deterioration in the primary balance of 0.9 percentage point in 2019.
  - If the shock is perfectly anticipated, the deterioration is 0.8 percentage point in 2019.
- Debt ratio dynamics (over a five year period, relative to baseline):
  - Disinflationary shock (above-described downward 2 percentage point shock) → debt ratios permanently increase by 6–7 percentage points of GDP relative to baseline.
  - Positive (inflationary) shock → debt ratios temporarily decrease by 0–3 percentage points of GDP, depending on the degree of anticipation.
- Mechanisms amplifying debt effects:
  - Stock effect: the higher the initial debt levels, the more sensitive the debt ratio is to price variations.
  - Accumulation effect: repeated shocks build up through the debt accumulation process.
  - Interest bill effect: small and persistent, with opposite sign to the effect on the primary balance.

### Overall empirical conclusions and caveats
- General finding:
  - Disinflationary surprises yield a permanent increase in primary expenditure ratios and a positive and partially temporary effect on revenue ratios. Anticipation dampens these effects.
  - Inflationary surprises have, at most, a temporary effect on expenditures.
- Deflationary contexts:
  - Very low or negative inflation rates tend to adversely affect fiscal performance: debt-to-GDP ratios tend to mechanically increase, revenue intake is weakened, and expenditures are difficult to cut in nominal terms because of downward rigidities.
  - Recessionary deflations (“bad” deflations) have the strongest impact on debt; during such episodes, debt-to-GDP ratios tend to rise rapidly.
  - “Good” deflations (deflation with positive economic growth) tend to mute negative impacts on debt-to-GDP ratios.
- Limitations and data considerations:
  - Dataset covers 21 advanced economies over 150 years but contains few modern episodes of deflation; pre–World War I institutional settings differed (limited scope for intervention, tax regimes not capturing nominal GDP changes effectively).
  - Modern fiscal frameworks may be more vulnerable: sophisticated tax regimes capture macro variations directly, large shares of expenditure are indexed to inflation, and negative indexation may be politically infeasible.
- Important conditional factors:
  - Harm from deflation is greater when accompanied by economic recession or stagnation.
  - Distinguishing whether deflation originates from aggregate supply or aggregate demand shocks is crucial for appropriate policy response, though difficult in practice.
  - Revenue and expenditure composition and political economy considerations matter for country-specific assessments.
  - Expectations and surprises about inflation can compound fiscal outcomes.

### Policy implications and recommendations
- Policymakers should:
  - Recognize asymmetrical fiscal behavior in deflationary episodes and consider deliberate countercyclical support when appropriate, noting that this increases fiscal deficits and should not be passive.
  - Act rapidly in budget execution to address unexpected downward revisions of inflation to avoid unwanted ratchet effects on government expenditure.
  - Revisit sophisticated fiscal policy tools and rules in light of low inflation/deflation risk, including examining nominally expressed multiyear expenditure ceilings for downward revision when inflation is below expectations.
  - Consider increasing the share of inflation-indexed bonds in the government’s financing mix to alleviate negative effects of deflation on debt ratios.
  - Tailor approaches to country specifics, including political economy dimensions.

### Appendix A — Japan: empirical episode highlights (1990–2013)
- Primary expenditure and revenue movements:
  - Primary expenditure-to-GDP ratio rose steadily by 12 percent during 1990–2013 (reflecting rising social spending associated with ageing).
  - Revenue-to-GDP ratio declined by 4 percent during 1990–2002 (reflecting tax alleviation measures and financial sector impairment), returning to its initial level by 2013.
  - Effective tax rate estimated broadly stable given weak progressivity, except for tax policy changes.
- Debt dynamics:
  - Public debt rose from 69 percent of GDP in 1990 to 243 percent in 2013.
  - The sole denominator effect of inflation explains roughly one fifth of the debt increase between 1990 and 2013: 32 percentage points out of 174 percentage points.
  - Nominal interest rates fell during the deflationary period but remained above the nominal growth rate on average.
- Institutional note:
  - Downward indexation was applied late and only for a few years; pension indexation that would have called for absolute declines was not applied from 2000–02 and was suspended after 2004.

### Key statistics and numeric highlights (preserved exactly)
- Shock magnitude: 2 percentage point (downward or upward) inflation shock.
- Primary expenditure response (unexpected downward shock) → 1.4 percentage points by 2019.
- Primary expenditure response (unexpected upward shock) → 0.3 percentage point lower by 2019.
- Expected shock expenditure effect → 1.2 percent of GDP.
- Revenue response (downward shock) → 0.4 percent of GDP by 2019.
- Revenue response (upward shock) → around -0.4 percent of GDP by 2019.
- Primary balance deterioration from 2 percentage point decline in inflation in 2015–16 → 0.9 percentage point in 2019 (0.8 if perfectly anticipated).
- Debt ratio changes over five years (relative to baseline):
  - Disinflationary shock → permanently increase by 6–7 percentage points of GDP.
  - Inflationary shock → temporarily decrease by 0–3 percentage points of GDP.
- Japan 1990–2013:
  - Primary expenditure-to-GDP ratio rose by 12 percent during 1990–2013.
  - Revenue-to-GDP ratio declined by 4 percent during 1990–2002.
  - Debt-to-GDP rose from 69 percent in 1990 to 243 percent in 2013.
  - Denominator effect explains 32 percentage points out of 174 percentage points of the debt increase.

*Source: IMF working paper content provided in the PDF chapter/section.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp15176.pdf_