## _wp10189

## Source details

**Canonical URL:** [_wp10189](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10189.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10189.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10189.pdf.json)

---

### I. Introduction — context and research question
- IMF’s April 2010 World Economic Outlook projects negative output gaps through at least 2014 for all G-7 economies.
- Common drivers of prolonged slack: tight credit conditions, deleveraging in the private sector, and fiscal retrenchment.
- Competing views on inflation prospects during prolonged slack:
  - Risk of prolonged undershooting of official inflation targets and deflation (Krugman (2010); Wolf (2010)).
  - View that inflation will stabilize at target rates well before output returns to potential, or that inflation risks may be tilted to the upside (Meltzer (2010); Halligan (2010)).
- Rationale for empirical approach:
  - Significant theoretical ambiguity motivates an event-study tracing inflation during historical episodes of persistent large output gaps (PLOG).
  - Sample constructed to provide a broad perspective on inflation outcomes during PLOGs: 25 episodes in 14 countries over the last 40 years.

### II. Relation to the literature — theory, estimation challenges, and empirical strategy
- Theoretical framework:
  - New Keynesian Phillips Curve (NKPC) formulation used: π_t = ω_f E_t(π_{t+1}) + ω_b π_{t-1} + ζ_t(mct), with mct mapping approximately linearly to the output gap y_t − y_t^p under competitive labor markets.
  - NKPC implication emphasized: a large negative output gap should lower inflation via ζ·(y_t − y_t^p).
- Empirical controversies and pitfalls:
  - Typical single-equation GMM estimates often report a small/insignificant output gap coefficient θ and a flattened Phillips curve in recent decades.
  - Identification challenges: weak instruments in GMM; need for valid and strong predictors of future inflation.
  - Omitted dynamics, potential nonlinearities, asymmetries for large gaps, and omitted variables (commodity prices, exchange rates, cost of finance) can bias NKPC estimates.
  - Alternative proxies for real marginal cost (e.g., labor share) can change estimated slopes; evidence is mixed.
- Empirical strategy of this paper:
  - Event-study focusing on episodes with large negative output gaps to document descriptive inflation dynamics during PLOGs rather than relying on single-equation GMM across all gap magnitudes.

### III. Data sources, definitions, and sample composition
- Coverage and period:
  - Start sample: all 15 high-income OECD-DAC members with quarterly output gap estimates available from the OECD’s Economic Outlook database (includes all G-7 economies).
  - Maximum time period considered: 1970Q1–2007Q4 (nearly forty years up to the eve of the global financial crisis); some series start later in the 1970s; New Zealand as late as 1980Q1.
- Output gap data:
  - Obtained from the OECD Economic Outlook No. 86 (production-function-based potential output estimates using total factor productivity, capital services, and potential employment).
  - Note: OECD Economic Outlook No. 87 data released in June 2010 used only in Section VI for post-2007 analysis; main analysis uses No. 86 series.
- PLOG episode definition:
  - At least eight consecutive quarters of negative output gaps exceeding 1.5 percent in absolute terms.
- Sample composition and summary metrics:
  - 25 PLOG episodes in 14 countries drawn from the 15-country start sample (all but Norway feature at least one PLOG episode; several countries feature two episodes; Canada features three).
  - Episode lengths vary between 9 and 25 quarters.
  - Average episode duration described as "a little more than three years".
  - Average output gap during episodes: 3.3 percent.

### IV. Inflation dynamics during PLOG episodes — empirical findings
- General pattern of inflation:
  - Clear trend of disinflation during historical PLOG episodes.
  - Disinflation tends to start somewhat ahead of the identified PLOG episode.
  - Strongest decline in inflation typically occurs during the first 4–5 quarters of the episode, followed by gradual further easing or bottoming-out at lower rates.
- Cross-episode and within-episode patterns:
  - Disinflation continues through later stages in episodes that begin with relatively high inflation; bottoming-out starts earlier in low-inflation episodes.
  - Episode-level comparison (initial vs final year): all but two data points lie below the 45-degree line — inflation falls in the overwhelming majority of episodes.
  - The two atypical cases with slight inflation increases began from very low (in one case negative) initial inflation rates.
  - Relationship between initial and final inflation rates is roughly proportional: higher initial inflation yields greater absolute deceleration but similar relative declines.
- Illustrative magnitudes:
  - Cross-episode representative aggregate: 4.4 percentage points, or 22 percent of the initial inflation rate.
  - Australia episode: inflation fell from just below 7 percent to 2 percent over three years, implying average annual disinflation of 1.6 percent, or about 23 percent of the initial inflation rate.
  - Disinflation is most rapid at initial inflation rates between 4 and 6 percent; regression evidence shows disinflation tapers off if initial inflation is already in the very low single digits.
  - Two outlier episodes: Sweden early 1990s and Japan 2000s featured slight inflation increases from very low starting rates.

### V. Explanations for disinflation tapering at very low inflation rates
- Candidate mechanisms:
  - Well-anchored inflation expectations and reduced price adjustment frequency among price-setters.
  - Downward nominal rigidities in wages and prices preventing large negative nominal adjustments, keeping inflation at low positive rates.
  - Globalization and increased import competition reducing domestic inflation sensitivity to spare capacity.
- Statistical factor:
  - Mean reversion: very low (or very high) inflation observations may contain temporary shocks that bias subsequent movement back toward a normal rate; contributes to curvature observed in the cross-episode trend.
- Assessment:
  - Mean reversion alone cannot fully account for increased stickiness at low inflation in post-1990 episodes; likely joint roles for anchored expectations and downward nominal rigidities.

### VI. Dynamics of key covariates during PLOG episodes
- Variables examined: unemployment rate, nominal wage growth, real unit labor cost (ULC), US$ oil price (WTI), nominal effective exchange rate (NEER), broad money growth (M3/M2 variants), real policy interest rate (nominal policy rate deflated by contemporaneous quarterly CPI, saar).
- Stylized within-episode patterns:
  - Unemployment: rises during first one to two years and then stabilizes slowly.
  - Nominal wage growth and real ULC: fall (after a few quarters).
  - Oil prices (WTI): on average fall somewhat (offsetting typical pre-episode rise).
  - NEER: no uniform trend; average episode features moderate currency depreciation.
  - Broad money growth: no clear average trend; roughly stable on average.
  - Real policy rate: tends to ease or stabilize after episode onset.
- Bivariate relationships with disinflation (degree defined as annualized decline relative to pre-episode inflation; alternative measure: share of quarters with declining four-quarter inflation):
  - Mild evidence for cross-sectional Phillips-curve: larger average output gap (absolute) coincides with greater disinflation.
  - No prima facie evidence for “speed limits”: change in output gap or real GDP growth during episode does not systematically limit disinflation.
  - Weaker labor markets (levels and changes in unemployment) associated with stronger disinflation; falling inflation coincides with slowing nominal wage growth.
  - Change in real ULC shows a counterintuitive sign: declines in real ULC tend to be more moderate in episodes with greater disinflation (mechanical effect where faster disinflation raises real ULC).
  - Appreciating exchange rate (NEER increase indicating appreciation) systematically associated with greater disinflation.
  - Changes in US$ oil price and growth rate of broad money appear largely unrelated to the degree of disinflation.
  - Real policy rate correlations are inconclusive, possibly reflecting two-way causation.
- Quarter-level contrasts:
  - "Unusual" quarters (quarterly CPI inflation rising, seasonally adjusted) systematically feature exchange rate depreciation and/or oil price increases.
  - Nominal ULC growth accelerates modestly in quarters of rising inflation relative to quarters of falling inflation; other variables show no statistically significant quarter-level differences.

### VII. Implications for the current inflation outlook (ongoing PLOG episodes post-2007)
- Identification of ongoing episodes:
  - OECD Economic Outlook No. 87 identifies ongoing PLOG episodes for many advanced economies beginning late 2008; several current episodes length listed as at least ≥12–14 quarters in the source.
- Comparisons: current episodes vs historical medians
  - Depth: latest downturn opened up larger negative output gaps than most historical precedents.
  - Headline inflation: fell very sharply in early quarters but swung back somewhat; core CPI inflation fell more gradually (median core fell from 2 percent pre-episode to 1¼ percent in most recent four quarters).
  - Unemployment: risen in line with historical patterns but may be plateauing earlier and at lower rates in some countries; nominal wage growth has come down from pre-episode rates.
  - Real ULC: unusually rose in the early part of the current downturn (chiefly reflecting labor hoarding and falling CPI).
  - NEER: wide variety across episodes; on average roughly unchanged.
  - Broad money growth: declined sharply across the board in current episodes (a departure from historical precedents).
- Interpretation and risks:
  - Deep downturn implies persistent disinflationary pressures historically associated with PLOGs; historical barrier at very low inflation rates may be tested in severely affected economies.
  - Downward nominal rigidities and anchored inflation expectations likely to prevent widespread lasting deflation absent exceptionally severe shocks.
  - Cross-sectional relationship: unemployment correlates with disinflation over the first year of recent episodes.
  - Tail risk: unconventional monetary policies and large central bank balance sheet expansions could raise the risk of credibility loss and a self-reinforcing currency crisis — characterized as low-probability but high-impact. Broad money growth has remained subdued and medium-term inflation expectations have not shown signs of unhinging in available evidence.

### VIII. Key policy-relevant conclusions and caveats
- Main conclusions:
  - Persistent large output gaps (PLOG) in advanced economies have historically led to significant disinflation, roughly proportional to initial inflation.
  - Disinflation tends to taper off at very low inflation rates; likely causes include downward nominal rigidities and well-anchored expectations.
  - Labor market slack and exchange rate appreciation are important contributors to disinflation; oil price declines support disinflation in several cases.
  - Little cross-sectional support in this sample for “speed limits” (faster recoveries raising inflation) or for a close link between broad money growth and disinflation.
- Policy implications:
  - Limited upside inflation risk where ample spare capacity is expected to persist; further disinflation may be constrained once inflation is very low.
  - Projections and policy should account for uncertainty in real-time output gap measurement.
  - Historical patterns should inform but not mechanically determine current policy given potential discontinuities (e.g., fiscal strains, central bank credibility shifts).
- Two emphasized caveats:
  - Analysis is conditional on accurate assessment of spare capacity; real-time output gap estimates can be subject to large ex-post revisions.
  - Historical relationships may break down if sharp structural discontinuities occur (e.g., loss of central bank credibility leading to unanchored inflation expectations).

*Source: _wp10189 (excerpted content provided).*

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

### _wp10189 - References .............................................................................................................

### I. Introduction — context and research question
- Most advanced economies are projected to recover only slowly from the global financial crisis; the IMF’s April 2010 World Economic Outlook projects negative output gaps through at least 2014 for all G-7 economies.
- Common drivers of prolonged slack include tight credit conditions, deleveraging in the private sector, and fiscal retrenchment.
- Competing views on inflation prospects:
  - Concerns about prolonged undershooting of official inflation targets and deflation risks (Krugman (2010); Wolf (2010)).
  - Views that inflation will stabilize at target rates well before output returns to potential, or that inflation risks are tilted to the upside (Meltzer (2010); Halligan (2010)).
- Reasons why large negative output gaps might exert only a muted downward pull on inflation:
  - Better-anchored inflation expectations and a flatter Phillips curve since the early 1990s.
  - Offsetting increases in non-labor input costs (commodity prices, depreciating currency).
  - Short-term “speed limits” to resource reallocation (Dale (2009); Dwyer et al. (2010)).
  - High fiscal deficits combined with unconventional monetary policy potentially eroding central bank credibility.
- Empirical approach motive:
  - Significant uncertainty and theoretical ambiguity motivate an empirical event-study approach tracing inflation during historical episodes of persistent large output gaps (PLOG).
  - Sample comprises 25 episodes in 14 countries over the last 40 years to provide a broad perspective on inflation outcomes during PLOGs.

### II. Relation to the literature — theory, estimation challenges, and empirical strategy
- New Keynesian Phillips Curve (NKPC) formulation (symbols preserved):
  - π_t denotes inflation between periods t-1 and t; mct denotes deviation of real marginal cost from steady-state.
  - NKPC (as presented):
    - General form: π_t = ω_f E_t(π_{t+1}) + ω_b π_{t-1} + ζ_t(mct) , with parameters ω_f, ω_b, and ζ (as in the text).
    - With competitive labor markets, mct maps approximately linearly to the output gap y_t − y_t^p (see equation (2) in the source).
  - Key NKPC prediction: a large negative output gap should lower inflation relative to its “normal” path (focus on ζ·(y_t − y_t^p)).
- Typical empirical findings and controversies:
  - NKPC estimation via single-equation GMM often reports a very small and frequently insignificant output gap coefficient θ; reported flattening of the Phillips curve in recent decades (cited studies: Galí and Gertler (1999); Sbordone (2002); Galí et al. (2005); Kleibergen and Mavroeidis (2008); Fuhrer and Moore (1995); Fuhrer (1997); Rudd and Whelan (2005a)).
  - Explanations for apparent flattening: better-anchored inflation expectations (Laxton and N’Diaye (2002); Roberts (2006); Mishkin (2007)) and globalization effects (IMF (2006); Borio and Filardo (2007)).
- Methodological pitfalls highlighted:
  - Weak instruments problem in GMM identification (Ma (2002); Mavroeidis (2005)); instruments must be valid and strong predictors of future inflation.
  - Omitted dynamics and possible nonlinearities: disproportionate effects when output gaps are large; asymmetry between positive and negative gaps; need for richer lag structures and allowance for drift in mean inflation (Ascari (2004); Laxton et al. (1995); Turner (1995)).
  - Omitted variable bias from drivers of real marginal cost other than spare capacity (commodity prices, exchange rates, cost of finance; Batini et al. (2005)).
  - Alternative proxies for real marginal cost (e.g., labor share in GNI) can yield larger NKPC slope coefficients (Galí and Gertler (1999); Woodford (2001)), though evidence is mixed (Rudd and Whelan (2005b)).
- Empirical strategy of this paper:
  - Event-study of 25 historical PLOG episodes to provide descriptive evidence on inflation dynamics during protracted negative output gap episodes.
  - Focus limited to the relevant segment of the Phillips curve (large negative gaps) and to broader trends over drawn-out episodes rather than short-term dynamics.

### III. Data sources, definitions, and sample composition
- Country coverage and period:
  - Start sample: all 15 high-income OECD-DAC members with quarterly output gap estimates available from the OECD’s Economic Outlook database (includes all G-7 economies).
  - Maximum time period considered: 1970Q1–2007Q4 (nearly forty years up to the eve of the global financial crisis); some series start later in the 1970s, New Zealand as late as 1980Q1.
- Output gap data:
  - Obtained from the OECD Economic Outlook No. 86 (production-function-based potential output estimates using total factor productivity, capital services, and potential employment; see Beffy et al. (2006)).
  - Note: OECD Economic Outlook No. 87 data were released in June 2010; the June 2010 dataset is used in Section VI of the source for post-2007 analysis, but the main analysis uses the No. 86 series.
- Definition of a PLOG episode:
  - At least eight consecutive quarters of negative output gaps exceeding 1.5 percent in absolute terms.
- Sample composition and summary metrics:
  - Among the 15 countries, all but Norway feature at least one PLOG episode; several countries feature two episodes; Canada features three.
  - Episode lengths vary between 9 and 25 quarters.
  - On average, a PLOG episode lasts "a little more than three years".
  - Average output gap during episodes: 3.3 percent (indicating severe demand shortfall).
- Stylized pattern of real GDP during episodes:
  - Figure 1 (as referenced) displays distribution of output gap paths focusing on the first nine quarters of each episode (t-0 denotes first quarter of each episode).
  - GDP-level chart (Figure 2) shows episodes were not periods of absolute decline; recessions are usually already underway when the PLOG episode begins; GDP fall tends to end around the second or third quarter of the episode and is followed by a rebound as the economy gradually closes the output gap.

### IV. Inflation dynamics during PLOG episodes — empirical findings
- General pattern:
  - Clear trend of disinflation during historical PLOG episodes.
  - Disinflation tends to start somewhat ahead of the identified PLOG episode.
  - The strongest decline in inflation occurs, on average, during the first 4–5 quarters of the episode, followed by gradual further easing or bottoming-out at lower rates.
- Distributional features:
  - Disinflation continues through later stages of PLOG episodes in cases that begin with relatively high inflation; bottoming-out starts earlier in low-inflation episodes.
  - Event windows examined:
    - First nine quarters of each episode (Figure 3, CPI inflation, percent, seasonally adjusted at an annual rate; X-axis t-4 … t+8 with t-0 the first quarter of each episode).
    - Final nine quarters of each episode (Figure 4, CPI inflation in later stages, percent, saar; X-axis T-8 … T-0 with T-0 the final quarter of each episode).
- Episode-level comparison (initial vs final year of episode):
  - Figure 5 compares CPI inflation in the year prior to the PLOG episode and in the final year of the episode for each episode listed in Table 2.
  - Evidence: all but two data points lie below the 45-degree line — inflation falls in the overwhelming majority of episodes.
  - The two atypical cases where inflation rose experienced negligible increases occurring from very low (in one case negative) initial inflation rates.
  - Relationship between initial and final inflation rates is roughly proportional: countries with high initial inflation exhibit greater absolute deceleration, but similar relative declines.
  - Annualized changes (percentage points per annum) between the final pre-episode year and the final year of the episode show similar relative declines across countries (regression line in Figure 6 referenced).
- Illustrative comparison (textual example as presented in the source):
  - U.K. early 1980s episode: inflation dropped from 20 percent to less than 5 percent over three-and-a-half years, implying average annual disinflation of (text breaks here in source).
- Summary conclusion from observed episodes:
  - Historical PLOG episodes in the sample predominantly involve clear disinflation; persistent large negative output gaps have typically been associated with substantial declines in CPI inflation over the course of the episode.

*Source: Text and figures excerpted from the source PDF content provided.*

### 4.4 percentage points, or 22 percent of the initial inflation rate. The Australian episode, in

### _wp10189 - 4.4 percentage points, or 22 percent of the initial inflation rate. The Australian episode, in

### Extent and speed of disinflation during PLOG episodes
- Disinflation across episodes is roughly proportional to initial inflation; example magnitudes:
  - Australia episode: inflation fell from just below 7 percent to 2 percent over three years, implying average annual disinflation of 1.6 percent, or about 23 percent of the initial inflation rate.
  - Representative aggregate: 4.4 percentage points, or 22 percent of the initial inflation rate (context: cross-episode measure cited at start of section).
- Figure-based findings:
  - Disinflation dominated during PLOG episodes, with greatest relative speed at inflation rates between 4 and 6 percent (Figure 7).
  - Regression line in Figure 6 crosses the x-axis at a positive value: disinflation tapers off if initial inflation is already in the very low single digits.
  - Two outlier episodes (Sweden early 1990s; Japan 2000s) featured slight increases in inflation from very low initial rates.

### Explanations for disinflation tapering at very low inflation rates
- Three economic hypotheses:
  - Enhanced central bank credibility and well-anchored inflation expectations: low steady-state inflation and less frequent price adjustments by price-setters (Ball et al. (1988) cited).
  - Downward nominal rigidities (wages and prices): as average inflation approaches zero, aversion to outright nominal cuts keeps inflation hovering at low positive rates (Akerlof et al. (1996); Benigno and Ricci (2010)).
  - Globalization effects: increased international competition and cheaper imports may have reduced domestic sensitivity of inflation to spare capacity, contributing to secular decline in inflation.
- Statistical argument:
  - Mean reversion: observations with the lowest/highest inflation rates are likelier to contain temporary shocks that push inflation back toward a “normal” rate; contributes to the curl of the polynomial trend in Figure 7.
- Evidence from subsamples/time periods:
  - Pre-1990 subsample: 14 observations; post-1990 subsample: 10 observations (Denmark 1989–92 excluded).
  - Post-1990 episodes feature, on average, lower initial inflation rates and less absolute disinflation, with sharp early falls followed by bottoming out (Figures 8, 9, 10).
- Assessment:
  - Mean reversion cannot fully explain increased stickiness at low rates in post-1990 episodes; likely roles for well-anchored expectations and downward nominal rigidities.

### Dynamics of key covariates during PLOG episodes
- Variables examined: unemployment rate, nominal wage growth, real unit labor cost (ULC), US$ oil price (WTI), nominal effective exchange rate (NEER), broad money growth (M3/M2 variants), real policy interest rate (nominal policy rate deflated by contemporaneous quarterly CPI, saar).
- Stylized patterns (Figure 11):
  - Unemployment rises during first one to two years and then stabilizes slowly.
  - Nominal wage growth and real ULC fall (after a few quarters).
  - Oil prices on average fall somewhat (offsetting typical pre-episode rise).
  - NEER shows no uniform trend; average episode features moderate currency depreciation.
  - Broad money growth shows no clear trend, remaining roughly stable on average.
  - Real policy rate tends to ease or stabilize after episode onset.
- Bivariate relationships with disinflation (Figures 12 and 13; degree of disinflation defined as annualized decline relative to pre-episode inflation; alternative: share of quarters with declining four-quarter inflation):
  - Mild evidence for cross-sectional Phillips-curve: larger average output gap (absolute) coincides with greater disinflation.
  - No prima facie evidence for “speed limits”: change in output gap or real GDP growth during episode does not systematically limit disinflation.
  - Weaker labor markets (levels and changes in unemployment) are associated with stronger disinflation; falling inflation coincides with slowing nominal wage growth.
  - Change in real ULC during episodes has counterintuitive sign: declines in real ULC tend to be more moderate in episodes with greater disinflation (mechanical/denominator effect where faster disinflation raises real ULC).
  - Appreciating exchange rate (NEER increase indicates appreciation) systematically associated with greater disinflation.
  - Changes in US$ oil price and growth rate of broad money appear largely unrelated to degree of disinflation.
  - Real policy rate correlations are inconclusive (mild positive with one measure, mild negative with another), possibly reflecting two-way causation.
- Unusual vs. typical quarters (Figure 14; unusual = quarter with rising quarterly CPI inflation sa):
  - Unusual quarters systematically feature exchange rate depreciation and/or oil price increases.
  - Statistically significant yet small link: nominal ULC growth accelerates in quarters of rising inflation relative to quarters of falling inflation (difference small in economic terms).
  - Other variables show no statistically significant difference between unusual and typical quarters.

### Implications for the current inflation outlook (ongoing PLOG episodes post-2007)
- Identification of ongoing episodes:
  - OECD Economic Outlook No. 87 identifies ongoing PLOG episodes for many advanced economies beginning late 2008; Table 3 lists countries and lengths (e.g., Australia 2008:4-2011:4 (at least) ≥13 quarters; multiple countries ≥12–14 quarters).
- Current episodes vs. historical medians (Figure 15):
  - Depth: latest downturn opened up larger negative output gaps than most historical precedents.
  - Headline inflation: fell very sharply in early quarters but swung back somewhat; core CPI inflation fell more gradually (median core fell from 2 percent pre-episode to 1¼ percent in most recent four quarters).
  - Unemployment: risen in line with historical patterns but may be plateauing earlier and at lower rates in some countries; nominal wage growth has come down from pre-episode rates.
  - Real ULC: unusually rose in early part of current downturn (chiefly reflecting labor hoarding and falling CPI).
  - NEER: wide variety across episodes; on average roughly unchanged.
  - Broad money growth: declined sharply across the board in current episodes (departure from historical precedents).
- Interpretation and risks:
  - Deep downturn argues for persistent disinflationary pressures; historical barrier at very low inflation rates may be tested in severely affected economies.
  - Downward nominal rigidities and anchored inflation expectations likely to prevent widespread lasting deflation absent exceptionally severe shocks.
  - Cross-sectional relationship: unemployment correlates with disinflation over the first year of recent episodes (Figure 17).
  - Tail risk concern: unconventional monetary policies and large central bank balance sheet expansions could raise the risk of a credibility loss and a self-reinforcing currency crisis—this is a low-probability but high-impact scenario. Broad money growth has remained subdued and medium-term inflation expectations have not shown signs of unhinging in available evidence.

### Key policy-relevant conclusions and caveats
- Main conclusions:
  - Persistent large output gaps (PLOG) in advanced economies have historically led to significant disinflation, roughly proportional to initial inflation.
  - Disinflation tends to taper off at very low inflation rates; likely causes include downward nominal rigidities and well-anchored expectations.
  - Labor market slack and exchange rate appreciation are important contributors to disinflation; oil price declines support disinflation in several cases.
  - Little cross-sectional support in this sample for “speed limits” (faster recoveries raising inflation) or for a close link between broad money growth and disinflation.
- Implications for current policy and outlook:
  - Little upside inflation risk where ample spare capacity is expected to persist; but further disinflation may be limited once inflation is very low.
  - Projections and policy should account for uncertainty in real-time output gap measurement.
  - Historical patterns should inform but not mechanically determine current policy given potential discontinuities (e.g., fiscal strains, central bank credibility shifts).
- Two caveats emphasized:
  - Analysis conditional on accurate assessment of spare capacity; real-time output gap estimates can be subject to large ex-post revisions.
  - Historical relationships may not hold if sharp structural discontinuities occur (e.g., loss of central bank credibility leading to unanchored inflation expectations).

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10189.pdf*

### References

### _wp10189 - References

### Journal articles and book chapters
- Akerlof, G., W. Dickens, and G. Perry (1996): “The Macroeconomics of Low Inflation”, Brookings Papers on Economic Activity, No. 1, 1-59.  
- Ascari, G. (2004): “Staggered Prices and Trend Inflation: Some Nuisances.” Review of Economic Dynamics, 7 (3), 642–667.  
- Ball, L., G. Mankiw, and D. Romer (1988): “The New Keynesian Economics and the Output–Inflation Trade-Off”, Brookings Papers on Economic Activity, 1, 1–82.  
- Batini, N., B. Jackson, and S. Nickell (2005): “An Open-Economy New Keynesian Phillips Curve for the U.K.”, Journal of Monetary Economics 52, 1061–1071.  
- Bean, C. (2007): “Globalisation and Inflation”, World Economics, 8, 1, 57–73.  
- Benigno, P. and L. Ricci (2010): “The Inflation-Output Trade-Off with Downward Wage Rigidities”, NBER Working Paper No. 15672.  
- Borio, C. and P. Disyatat (2009): “Unconventional Monetary Policies: An Appraisal”, BIS Working Paper No. 292.  
- Borio, C. and A. Filardo (2007): “Globalisation and Inflation: New Cross-Country Evidence on the Global Determinants of Domestic Inflation”, BIS Working Paper No. 227.  
- De Veirman, E. (2009): “What Makes the Output-Inflation Trade-Off Change? The Absence of Accelerating Deflation in Japan”, Journal of Money, Credit, and Banking 41 (6), 1117–1140.  
- Dwyer, A., K. Lam, and A. Gurney (2010):”Inflation and the Output Gap in the U.K.”, HM Treasury Economic Working Paper No. 6.  
- Fuhrer, J.C. and G.R. Moore (1995): “Inflation persistence”, Quarterly Journal of Economics 440, 127–159.  
- Fuhrer, J.C. (1997): The (Un)importance of Forward-Looking Behavior in Price Specifications”, Journal of Money, Credit, and Banking 29 (3), 338–350.  
- Galí, J. and M. Gertler (1999): “Inflation Dynamics: A Structural Econometric Analysis”, Journal of Monetary Economics 44, 195–222.  
- Galí, J., M. Gertler, and D. Lopez-Salido (2005): "Robustness of the Estimates of the Hybrid New Keynesian Phillips Curve", Journal of Monetary Economics, 52 (6), 1107–1118.  
- Keister, T., A. Martin, and J. McAndrews (2008): “Divorcing Money from Monetary Policy,” Federal Reserve Bank of New York Economic Policy Review, 14 (2).  
- Kleibergen, F. and S. Mavroeidis (2009): “Weak Instrument Robust Tests in GMM and the New Keynesian Phillips Curve”, Journal of Business and Economic Statistics, 27, 293–311.  
- Ma, A. (2002): “GMM Estimation of the New Keynesian Phillips Curve”, Economics Letters, 76, 411–417.  
- Mavroeidis, S. (2005): “Identification Issues in Forward-Looking Models Estimated by GMM With an Application to the Phillips Curve”, Journal of Money Credit and Banking, 37 (3), 421–449.  
- Mishkin, F. (2007): “Inflation Dynamics”, NBER Working Paper No. 13147.  
- Nason, J. M. and G.W. Smith (2008): “The New Keynesian Phillips Curve: Lessons from Single-Equation Econometric Estimation”, Federal Reserve Bank of Richmond Economic Quarterly, 94 (4), 361–395.  
- Orphanides, A., R. D. Porter, D. Reifschneider, R. Tetlow, and F. Finan (2000): ”Errors in the Measurement of the Output Gap and the Design of Monetary Policy”, Journal of Economics and Business, 52, 117–141.  
- Roberts, J. M. (2006): "Monetary Policy and Inflation Dynamics", International Journal of Central Banking, 2, 193–230.  
- Rotemberg, J. and M. Woodford (1999): “The Cyclical Behavior of Prices and Costs”, Handbook of Macroeconomics, edited by J. Taylor and M. Woodford, 1051–1135, Amsterdam: North-Holland.  
- Rudd, J., and K. Whelan (2005a): “New Tests of the New Keynesian Phillips Curve”, Journal of Monetary Economics, 52 (6), 1167–1181.  
- Rudd, J., and K. Whelan (2005b): “Does Labor’s Share Drive Inflation?” Journal of Money, Credit, and Banking, 37, 2, 297–312.  
- Sbordone, A. M. (2002): “Prices and Unit Labor Costs: A New Test of Price Stickiness”, Journal of Monetary Economics, 49, 265–292.  
- Turner, D. (1995): “Speed Limit and Asymmetric Inflation Effects from the Output Gap in the Major Seven Economies”, OECD Economic Studies, 24, 57–87.  
- Walsh, C. (2003):”Minding the Speed Limit”, Federal Reserve Bank of San Francisco Economic Letter, 2003–2014.  
- Woodford, M. (2001): “The Taylor Rule and Optimal Monetary Policy”, American Economic Review, 91, 232–237.

### Working papers, reports, and policy notes
- Beffy, P.-O., P. Ollivaud, and F. Sédillot (2006): “New OECD Methods for Supply-Side and Medium-Term Assessments: A Capital Services Approach”, OECD Economics Department Working Paper No. 482.  
- Gagnon, J. E., M. Raskin, J. Remache, and B. P. Sack (2010): “Large-Scale Asset Purchases by the Federal Reserve: Did They Work?” FRB of New York Staff Report No. 441.  
- Laxton, D., G. Meredith, and D. Rose (1995): “Asymmetric Effects of Economic Activity on Inflation: Evidence and Policy Implications”, IMF Staff Papers, 42, 2, 344–374.  
- Laxton, D. and P. N’Diaye (2002): “Monetary Policy and the Unemployment-Inflation Tradeoff: Some Evidence from 17 Industrial Countries”, IMF Working Paper No. 02/222.  
- Meier, A. (2009): “Panacea, Curse, or Nonevent? Unconventional Monetary Policy in the United Kingdom”, IMF Working Paper 09/163.  
- Borio, C. and P. Disyatat (2009): “Unconventional Monetary Policies: An Appraisal”, BIS Working Paper No. 292.  
- Borio, C. and A. Filardo (2007): “Globalisation and Inflation: New Cross-Country Evidence on the Global Determinants of Domestic Inflation”, BIS Working Paper No. 227.

### Op-eds, blogs, and remarks
- Dale, S. (2009): “2009: A Review of the Economic Year”, Remarks at the Chairman’s Annual Breakfast, Essex Institute of Directors, Billericay, December 2.  
- Halligan, L. (2010): “Our Politicians Have Been Happy to Promote the Deflationary Myth”, Daily Telegraph Op-ed, April 24.  
- Krugman, P. (2010): “Crying Fire! Fire! In Noah’s Flood”, New York Times blog posted May 21 on http://krugman.blogs.nytimes.com/2010/05/21/crying-fire-fire-in-noahs-flood-3/.  
- Meltzer, A. (2010): “The Fed's Anti-Inflation Exit Strategy Will Fail”, Wall Street Journal Op-ed, January 27.  
- Wolf, M. (2010): “Fear Must not Blind Us to Deflation’s Dangers”, Financial Times, June 8.

### Institutional publications and other sources
- IMF (2006), “How Has Globalization Affected Inflation?” IMF World Economic Outlook, April, Chapter 3.  
- Keister, T., A. Martin, and J. McAndrews (2008): “Divorcing Money from Monetary Policy,” Federal Reserve Bank of New York Economic Policy Review, 14 (2).  
- Nason, J. M. and G.W. Smith (2008): “The New Keynesian Phillips Curve: Lessons from Single-Equation Econometric Estimation”, Federal Reserve Bank of Richmond Economic Quarterly, 94 (4), 361–395.  

*Source: _wp10189 - References*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10189.pdf_