## INTRODUCTION — IMF Working Paper wp18147 (INTRODUCTION)

## Source details

**Canonical URL:** [INTRODUCTION — IMF Working Paper wp18147 (INTRODUCTION)](https://www.imf.org/-/media/files/publications/wp/2018/wp18147.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2018/wp18147.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2018/wp18147.pdf.json)

---

### Overview and aims
- Focus on inflation dynamics in the five founding members of ASEAN: Indonesia, Malaysia, the Philippines, Singapore and Thailand.
- Twofold aims:
  - Provide evidence on how inflation dynamics evolved in ASEAN-5 economies over the last three decades.
  - Explore the extent to which changes in inflation dynamics can be attributed to the improvement of monetary policy frameworks.

### Empirical methodology and data
- Benchmark model: hybrid New Keynesian Phillips Curve (NKPC) extended to open-economy context with time-varying parameters.
- Benchmark Phillips Curve (equation (1)):
  - π_t = β_{1,t} π^*_t + (1−β_{1,t}) π_{t−1}^{MA4} + β_{2,t} y^*_{t−1} + β_{3,t} π^{IM}_t + ε_t
  - Definitions preserved from source:
    - π_t: headline consumer price index inflation.
    - π^*_t: long-run inflation expectations.
    - π_{t−1}^{MA4}: moving average of inflation over the previous four quarters.
    - y^*_{t−1}: economic slack measured as the output gap.
    - π^{IM}_t: inflation of imported goods and services.
    - ε_t: measurement error, Gaussian white noise.
  - Coefficient interpretations:
    - β_{1,t}: measures forward-looking component; (1−β_{1,t}) captures persistence.
    - β_{2,t}: slope of the Phillips Curve (impact of cyclical activity).
    - β_{3,t}: effect of import price inflation.
- Data:
  - Quarterly data from 1995Q1 until 2016Q4 for ASEAN-5; Indonesia and Malaysia sample starts in 1998Q1.
  - Headline CPI and real GDP from WEO database; import price indices from HAVER database.
  - Import price and headline CPI included as year-on-year inflation rates.
  - Output gap computed using HP-filter.
  - Benchmark long-term inflation expectations: trend estimates from Garcia and Poon (2018).
  - For contributions and robustness: crude oil price and nominal exchange rates from HAVER; Consensus long-run inflation expectations.
- Estimation approach:
  - Kalman filter with Gauss-Newton optimization with Marquardt step.
  - Starting values from OLS over a ten-year rolling window.
  - Country-specific variance ratios introduced based on signal-to-noise ratios from rolling window estimation.
  - Decomposition of import price inflation into non-oil-import inflation and oil price inflation for 2014-16 disinflationary episode using regression analysis in line with IMF (2013, 2016).

### Summary of main findings — region-wide
- Forward-looking dynamics and shares (median or averages as in source):
  - Long-term expectations (forward-looking dynamics): explain on average around 60% of median inflation in the region.
  - Economic slack: explain on average 9% of median inflation.
  - Non-oil-import inflation: explain on average 12% of median inflation.
  - Oil price inflation: explain on average 7% of median inflation.
  - Estimation residual: broadly in line with 9%.
- Evolution of forward-looking dynamics (median shares by period):
  - 1996–2001: forward-looking dynamics averaged 43%.
  - Thereafter (post-AFC): increased to around 66%.
  - During the GFC: declined temporarily to 57%.
  - Recent disinflation period 2014–16: declined to 61%.
- Economic slack contributions:
  - 1995–2007: economic slack explained around 7% of headline inflation.
  - 2002–07 expansion: rose to 15%.
  - Thereafter declined to less than 5%, though higher relevance observed in recessionary episodes.
- Import price inflation contributions over three decades:
  - During the AFC: non-oil import and oil price inflation jointly were major drivers (almost 27%).
  - Thereafter: contributions declined to around 8% (non-oil) and 5% (oil) for most of the sample.
  - Recent disinflation episode 2014–16: non-oil-import and oil price inflation contributions rebounded to 15% and 18% of median headline inflation respectively.

### The role of forward-looking dynamics and central bank transparency
- Key mechanism:
  - In the hybrid NKPC, β_{1,t} determines importance of long-term expectations; (1−β_{1,t}) represents inflation persistence.
  - Trend inflation estimates (Garcia and Poon, 2018) incorporate survey-based expectations while allowing for deviations.
- Observed features:
  - Survey expectations are substantially more volatile than trend inflation estimates.
  - For Malaysia, Indonesia and Thailand, trend estimates lie below Consensus survey expectations throughout the sample.
- Central bank transparency (Dincer and Eichengreen (2014) DE index):
  - Starting scores between 2 and 4 index points in 1998.
  - By 2014, scores: Indonesia, Philippines and Thailand between 9 and 10; Malaysia 6; Singapore 5.
  - Maximum score of 15.
- Empirical relation:
  - Time-varying estimated forward-looking coefficient β_{1,t} and the contribution β_{1,t} π^*_t are positively correlated with each country’s DE transparency score.
  - Results support that greater transparency is associated with higher degree of forward-looking dynamics.

### Inflation dynamics and cyclical fluctuations (Phillips Curve slope)
- Slope and contributions:
  - β_{2,t} is key for link between economic slack and inflation.
  - Median contribution of output gap to inflation fluctuated significantly over three decades; more important ahead of the GFC; very limited role thereafter except in recessionary episodes.
- Phillips curve flattening:
  - All ASEAN-5 countries experienced a flattening of the Phillips curve in recent years.
  - Timing and magnitude vary by country.
  - ASEAN-5 flattening consistent with evidence for many advanced economies post-GFC.
- Policy implication:
  - Flattening helps explain relatively low inflation in ASEAN-5 in 2017 despite growth momentum.

### Inflation dynamics, non-oil-import and oil price inflation
- Combined contributions declined substantially after the AFC.
- Non-oil import inflation declined markedly since early 2000s; oil price inflation contribution rose after AFC.
- Historical ranges for oil price contributions (median contribution to headline inflation):
  - 1996–2001: oil price contributions lie between 0.30 percentage points and -0.49 percentage points.
  - 2002–08: contributions range increases to (1.41 -0.17 percentage points).
  - Recent disinflation: oil price inflation drove down median inflation by -0.46 percentage points in 2015 and -0.59 percentage points in 2016.
- Synthesis:
  - Long-run inflation expectations became the most important driver.
  - Output gap movements more limited in normal conditions but stronger during crises.
  - Import inflation contribution has become more limited over time, relatively stable since early 2000s.

### Country-specific evidence — key statistics and highlights

- Indonesia
  - Aggregate movements:
    - Indonesian inflation declined from an average of 8.5% before the GFC to around 5% after the GFC.
  - Example years and contributions:
    - In 2002: Indonesian inflation ≈ 12%; contributions: economic activity 3.5 percentage points; non-oil-import prices 2.1 percentage points; oil price inflation 0.8 percentage points; expectations 3.6 percentage points.
    - In 2015: Headline inflation 6.4%; contributions: expectations 4.9 percentage points; economic slack 0.05 percentage points; non-oil-import inflation 0.8 percentage points; oil price inflation -0.5 percentage points.
  - Time-varying coefficients and policy:
    - Contribution of forward-looking dynamics increased by 74% from 2001 (2.67 percentage points) to 2005 (4.7 percentage points).
    - Forward-looking coefficient rose from 0.32 in 2001 to 0.73 in 2007, stabilizing at around 0.7.
    - Trend inflation fell from 7.9% in 2001 to 5% from 2007 onwards.
    - Output gap contributions declined from a range of 4.2 to -1.7 percentage points in 2001-07 to 0.4 to -0.7 percentage points in 2008-16.
    - Phillips Curve slope declined from 0.49 in 2001 to 0.03 in 2010.
    - Coefficient of import prices: 0.13 in 2001 → 0.31 in 2006 → around 0.08 from 2008-16.
    - Contributions of non-oil-import and oil price inflation declined from averages of 2.4 and 1.5 percentage points between 2001-08 to 0.83 and 0.13 percentage points thereafter.

- Malaysia
  - Average headline inflation: 2.5% over sample.
  - Decomposition:
    - Forward-looking expectations: 2.08 percentage points.
    - Economic slack and import inflation combined: 0.2 percentage points (output gap 0.02 percentage points, non-oil-import 0.06 percentage points, oil 0.11 percentage points).
  - Changes over time:
    - Contribution of inflation expectations: 1.8 percentage points (2002-07) → 2.3 percentage points (2008-16).
    - Trend inflation stable around 2.5% throughout sample.
    - Forward-looking coefficient: 0.4 over 2002-05 → 0.8 over 2006-16.
    - Output decline during GFC contributed between 3.2 to -1.6 percentage points over 2008-10.
    - Slope parameter roughly 0.4 across sample, increases to 0.8 over 2008-10.
    - Contributions of import inflation range from 0.46 to -0.24 percentage points for non-oil-import and 0.37 to -0.26 percentage points for oil price inflation over entire sample.
    - Coefficient of import inflation increases from 0.006 over 2002-08 to 0.1 over 2009-16.

- Philippines
  - Forward-looking dynamics:
    - Forward-looking dynamics account for 72% of headline inflation on average (3.6 percentage points of 4.9% headline inflation) over entire sample.
    - 1996-2007: 64% of average headline inflation (3.64 percentage points of 5.8% headline inflation) explained by forward-looking dynamics.
    - Since GFC: share increased to 87% (3.4 percentage points of 3.9% headline inflation over 2008-16).
  - Trend and coefficients:
    - Long-run trend inflation declined from around 4% between 1995-2008 to 2.9% in 2016.
    - Coefficient on trend expectations stable around 0.4 across entire sample.
    - Coefficient of economic slack increased from 0.03 in 1996 to 0.81 in 2008.
    - Standard deviation of output gap declined from 1.33 over 1996-2009 to 0.74 from 2010 onwards.
    - Contribution of economic slack over entire sample ranges between 2.71 to -1.29 percentage points.
  - Import price contributions:
    - Over 1996-2002, on average 28% of headline inflation associated with non-oil-import inflation (1.7 percentage points of 6% average headline inflation).
    - Between 2003-16 relative contribution declined to 3% (0.1 percentage points of 4.4% average headline inflation).
    - Oil price contributions range from 1 to -0.6 percentage points.
    - Coefficient of import inflation rises from 0.1 in 1996 to 0.41 in 2006 and stabilizes around 0.23 thereafter.

- Singapore
  - Average headline inflation: 1.6% between 1996 and 2016.
  - Decomposition averages:
    - Forward-looking dynamics: 1.85 percentage points.
    - Economic slack: 0.16 percentage points.
    - Non-oil-import inflation: 0.47 percentage points.
    - Oil price inflation: -0.11 percentage points.
  - Volatility and disinflation:
    - Inflation fell from 4.7% in 2012 to 1% in 2014, and into negative territory in 2015 and 2016.
    - Contribution of inflation expectations increased from 0.2 percentage points in 1996 to an average of 2.2 percentage points between 2007 and 2016, explaining 88% of average headline inflation during that period.
    - Forward-looking coefficient rises from 0.19 in 1996 to 0.34 in 2016, with a drop to 0.05 in 2008 and rise to 0.39 in 2010.
  - Import pass-through and composition:
    - Average contribution of non-oil-import and oil price inflation: 0.59 percentage points between 1996 and 2006; 1.9 percentage points from 2007 onwards.
    - Import-inflation coefficient: 0.06 in 1996 → 0.15 in 2014 → decreasing trend thereafter.
    - Noted possible disinflationary factor: steady decline in property prices, reflected in a negative residual toward the end of sample.

- Thailand
  - Forward-looking dynamics:
    - Forward-looking dynamics explains more than half (53%) of headline inflation after the AFC versus 33% before the AFC.
    - Absolute contribution of forward-looking dynamics: 2.4 percentage points from 2001 to 2010, decreases to 2 percentage points thereafter.
  - Headline inflation trend:
    - Declined from an average of 6.5% before the AFC to 2.5% thereafter.
    - Episodes of deflation in 2008 and 2015 due to negative output gap and oil price pass-through from 2014 onwards.
  - Output gap and coefficients:
    - Average contribution during AFC: -2.13 percentage points.
    - Average contribution during GFC: -2.58 percentage points.
    - Average contribution since 2010: -0.08 percentage points.
    - Time-varying coefficient of economic slack: 0.03 in 1995 → 0.54 in 2008 → 0.3 in 2015.
  - Import-price contributions:
    - Absolute contributions (entire sample) of non-oil-import and oil price inflation average 0.54 and 0.75 percentage points, respectively.
    - Contribution of non-oil-import inflation declines from 1.44 percentage points during the AFC to 0.47 percentage points thereafter.
    - Coefficient of import inflation increases from 0.08 in 1996 to 0.2 in 2008 and remains stable thereafter.
  - Policy and communications:
    - BOT operated under an inflation targeting regime since 2000.
    - Change of target from core to headline in early 2015 coincided with intensifying disinflationary pressures.
    - Memorandum of Understanding requires MPC to explain missing the inflation target annually and issue a letter outlining the period and policy response.

### Robustness checks — data choices and model specifications
- General: qualitative findings robust to series of sensitivity analyses.
- Different measures of macroeconomic indicators:
  - Output-gap measures:
    - Benchmark HP-filter vs two-sided band-pass filter: differences quantitatively marginal; qualitative results similar.
    - Coefficients of output gap slightly higher during the GFC under alternative filter.
  - Inflation expectations measures:
    - Re-estimation using Consensus long-run inflation expectations yields qualitatively similar results.
    - Consensus survey expectations systematically higher than long-run trend estimates.
    - Singapore estimation results exclusively based on long-term Consensus Forecasts due to uncertainty in trend inflation estimation.
  - Import-price measures:
    - Re-estimation with import prices from WEO database yields qualitative implications in line with benchmark.
    - Combined contribution of oil and import price inflation quantitatively similar, weight shifts toward oil price inflation in that variation.
- Model specification checks:
  - Time-varying parameters vs constant parameters:
    - Ignoring time variation misses significant variation in inflation contributions; constant parameters can both under- and overestimate contributions.
    - Time-variation of the Phillips Curve slope most relevant during recessionary episodes.
    - Non-oil-import inflation tends to be overestimated under constant parameters.
  - Four-parameter model decomposing import inflation (non-oil-import and oil price inflation):
    - Median residuals rise to around 30% versus roughly 10% in benchmark.
    - Contribution of oil price inflation substantially altered; forward-looking dynamics contributions quantitatively lower than benchmark.
    - Country-specific coefficients of non-oil-imports systematically higher by roughly 0.05 but follow same dynamics.
    - Significant oil price inflation coefficients range from 0.005 to 0.028 across countries and time.
  - Models including exchange rate:
    - Augmenting model with nominal exchange rate shows limited pass-through to headline inflation.
    - Exchange-rate coefficient not significant across countries and time.
    - Median and country-specific contributions of exchange rate quantitatively very small.
    - Median and country-specific contributions of oil price inflation larger than in benchmark.

### Conclusions and policy implications
- Major conclusions:
  - Monetary policy frameworks have played a strong role in shaping inflation dynamics in ASEAN-5.
  - Inflation has become increasingly forward-looking (less persistent); inflation expectations are the most important component in all countries.
  - Forward-looking coefficient and contribution of inflation expectations positively related to central bank transparency (Dincer and Eichengreen (2014)).
  - Quantitative impact of cyclical fluctuations has gradually diminished; significant flattening of the Phillips curve since the GFC.
  - Increased exchange rate flexibility and economic integration may have reduced the effect of non-oil and oil import inflation on headline inflation.
- Remaining challenges and recommendations:
  - Strong disinflation and, in some cases, outright deflation in oil-importing economies over 2014-16 indicate supply-side shocks not fully offset by forward-looking component.
  - Thailand example: weakening of forward-looking dynamics and anchoring exacerbated deflationary outcome.
  - Policy recommendations highlighted in country discussions:
    - Continuous communication and policy actions to manage expectations and better align them to the inflation target.
    - Further guidance on envisaged actions to achieve the target over a given time horizon, critical assessment of past performance, detailed discussion of shocks, and elaboration of internal MPC debate (including dissenting views) can help guide private-sector expectations.
    - Announcing an inflation target over a longer horizon than one year may contribute to stronger anchoring of inflation expectations.

*IMF Working Paper — wp18147 (INTRODUCTION) — content unit as provided*

### Introduction ...........................................................................................................

### Introduction

### Empirical Methodology and Data
- Modeling Inflation Dynamics
- Data
- Estimation Approach
- (Section pages indicated: Introduction ........................................................................................................................... 5; Empirical Methodology and Data ........................................................................................ 7)

### III. What drives inflation dynamics in ASEAN-5 countries?
- A. Key Findings
- The Role of Forward-Looking Dynamics
- Inflation Dynamics and Cyclical Fluctuations
- Inflation Dynamics, Non-oil-import and Oil Price Inflation
- (Section marker: III. What drives inflation dynamics in ASEAN-5 countries? ................................................... 8)

### Country-specific evidence
- Indonesia
- Malaysia
- Philippines
- Singapore
- Thailand
- (Subsections with per-country headings and associated figure references)

### Robustness checks
- A. Different Measures of Macroeconomic Indicators
  - Measures of the Output Gap
  - Indicators of Inflation Expectations
  - Import Price Measures
- B. Model Specifications
  - Impact of Time-varying Parameters
  - Coefficients of Non-Oil-Import Price and Oil Price Inflation
  - Models including oil price inflation and exchange rate
- (Section pages indicated: Robustness checks.............................................................................................................. 30)

### Conclusions
- (Conclusions....................................................................................................................... 34)

### VI. References
- (VI. References......................................................................................................................... 35)

### Figures (listed)
- Figure 1: Relative Median Contribution of Inflation Drivers in ASEAN-5 countries ............. 9
- Figure 2: Long-term Inflation Expectations (Trend Inflation Estimates) ............................... 11
- Figure 3: Consensus Long-term Inflation Expectations ......................................................... 11
- Figure 4: ASEAN-5 CBs: Dincer-Eichengreen CB transparency index ................................ 14
- Figure 5: Correlation Between Forward-looking Dynamics and DE transparency: Coefficients ............................................................................................................................. 14
- Figure 6: Correlation Between Forward-looking Dynamics and DE transparency: contributions ........................................................................................................................... 14
- Figure 7: Median Contribution of Output Gap in ASEAN-5 countries .................................. 16
- Figure 8: Phillips curve Slope Estimates                                                                                  16
- Figure 9: Median Contribution of Import Inflation in ASEAN-5 Countries                           17
- Figure 10: Main Inflation Components: Indonesia ................................................................. 18
- Figure 11: Time-varying Coefficient Estimates: Indonesia .................................................... 19
- Figure 12: Main Inflation Components: Malaysia .................................................................. 21
- Figure 14: Main Inflation Components: Philippines .............................................................. 21
- Figure 15: Time-varying Coefficient Estimates: Philippines ................................................. 24
- Figure 16: Main Inflation Components: Singapore ................................................................ 24
- Figure 17: Time-varying Coefficient Estimates: Singapore ................................................... 26
- Figure 18: Main Inflation Components: Thailand .................................................................. 28
- Figure 20: Contributions resulting from time-varying parameters:Thailand .......................... 28

### Tables (listed)
- Table 1: Monetary policy Frameworks and Transparency in ASEAN-5 Countries ............... 13

*Source: wp18147 - Introduction ...........................................................................................................*

### INTRODUCTION

### INTRODUCTION

### Overview and Aims
- Inflation dynamics in the five founding members of ASEAN (Indonesia, Malaysia, the Philippines, Singapore and Thailand) have experienced substantial changes over the last 20 years.
- Twofold aim of the paper:
  - Provide evidence on how inflation dynamics evolved in ASEAN-5 economies over the last three decades.
  - Explore the extent to which changes in inflation dynamics can be attributed to the improvement of monetary policy frameworks.

### Summary of main findings
- Adoption of price stability objectives, higher central bank transparency and enhanced communication:
  - Improved anchoring of inflation expectations.
  - Led to stronger forward-looking dynamics (lower persistence) in inflation over the last two decades.
- Quantitative impact of cyclical fluctuations has gradually diminished over time.
- Significant flattening of the Phillips curve in ASEAN economies since the Global Financial Crisis (GFC).
- Exchange rate liberalization, economic development and technological progress contributed to reducing the effect of non-oil and oil import inflation on headline inflation.
- Strong disinflation and, in some cases, outright deflation in oil-importing economies (Singapore, Thailand) since 2014 indicate supply-side shocks have not been fully counterbalanced by the forward-looking component of the inflation process in ASEAN-5 yet.

*Further structure: Section II describes benchmark model and data; Section III discusses region-wide findings and Phillips curve flattening since the GFC; Section IV provides country-specific results; Section V robustness checks; Section VI concludes.*

---

### Empirical methodology and data
- Benchmark model: hybrid New Keynesian Phillips Curve (NKPC) extended to open-economy context with time-varying parameters.
- Benchmark Phillips Curve (equation (1)):
  - π_t = β_{1,t} π^*_t + (1−β_{1,t}) π_{t−1}^{MA4} + β_{2,t} y^*_{t−1} + β_{3,t} π^{IM}_t + ε_t
  - Definitions preserved as in source:
    - π_t: headline consumer price index inflation.
    - π^*_t: long-run inflation expectations.
    - π_{t−1}^{MA4}: moving average of inflation over the previous four quarters.
    - y^*_{t−1}: economic slack measured as the output gap.
    - π^{IM}_t: inflation of imported goods and services.
    - ε_t: measurement error, Gaussian white noise.
  - β_{1,t}: measures forward-looking component; (1−β_{1,t}) captures persistence.
  - β_{2,t}: slope of the Phillips Curve (impact of cyclical activity).
  - β_{3,t}: effect of import price inflation.
- Data:
  - Quarterly data from 1995Q1 until 2016Q4 for ASEAN-5; Indonesia and Malaysia sample starts in 1998Q1.
  - Headline CPI and real GDP from WEO database; import price indices from HAVER database.
  - Import price and headline CPI included as year-on-year inflation rates.
  - Output gap computed using HP-filter.
  - Benchmark long-term inflation expectations: trend estimates from Garcia and Poon (2018).
  - For contributions and robustness: crude oil price and nominal exchange rates from HAVER; Consensus long-run inflation expectations.
- Estimation approach:
  - Kalman filter with Gauss-Newton optimization with Marquardt step.
  - Starting values from OLS over a ten-year rolling window.
  - Country-specific variance ratios introduced based on signal-to-noise ratios from rolling window estimation.
  - Decomposition of import price inflation into non-oil-import inflation and oil price inflation for 2014-16 disinflationary episode using regression analysis in line with IMF (2013, 2016).

---

### What drives inflation dynamics in ASEAN-5? — Key findings
- Relative median contributions (across countries and over time) of drivers to headline inflation (median shares stated exactly as in source):
  - Long-term expectations (forward-looking dynamics): explain on average around 60% of median inflation in the region.
  - Economic slack: explain on average 9% of median inflation.
  - Non-oil-import inflation: explain on average 12% of median inflation.
  - Oil price inflation: explain on average 7% of median inflation.
  - Estimation residual: broadly in line with 9%.
- Evolution of forward-looking dynamics (median shares by period):
  - 1996–2001: forward-looking dynamics averaged 43%.
  - Thereafter (post-AFC): increased to around 66%.
  - During the GFC: declined temporarily to 57%.
  - Recent disinflation period 2014–16: declined to 61%.
- Economic slack contributions:
  - 1995–2007: economic slack explained around 7% of headline inflation.
  - 2002–07 expansion: rose to 15%.
  - Thereafter declined to less than 5%, though higher relevance observed in recessionary episodes.
- Import price inflation contributions over three decades:
  - During the AFC: non-oil import and oil price inflation jointly were major drivers (almost 27%).
  - Thereafter: contributions declined to around 8% (non-oil) and 5% (oil) for most of the sample.
  - Recent disinflation episode 2014–16: non-oil-import and oil price inflation contributions rebounded to 15% and 18% of median headline inflation respectively.

---

### The role of forward-looking dynamics
- Key insight: inflation expectations have become the most important driver of inflation dynamics across ASEAN-5.
- In the hybrid NKPC, β_{1,t} and π^*_t are crucial:
  - β_{1,t} determines importance of long-term expectations.
  - (1−β_{1,t}) represents inflation persistence.
- Trend inflation estimates used (Garcia and Poon, 2018) incorporate survey-based expectations while allowing for deviations.
- Observed features:
  - Survey expectations are substantially more volatile than trend inflation estimates.
  - For Malaysia, Indonesia and Thailand, trend estimates lie below Consensus survey expectations throughout the sample — pointing toward systematic deviations potentially explained by informational rigidities.

---

### Forward-looking inflation dynamics and central bank transparency
- Central bank transparency matters for aligning long-run trend expectations (π^*_t) with central bank targets and for inducing forward-looking pricing behavior.
- Historical evolution:
  - Before AFC: pegged exchange rate regimes dominated; exchange rate pressures followed capital flow reversals.
  - Since AFC: increased exchange rate flexibility and substantial improvements in operating frameworks, policy objectives and communication across ASEAN-5.
- Dincer and Eichengreen (2014) central bank transparency index (DE index) observations:
  - Starting scores between 2 and 4 index points in 1998.
  - By 2014, scores: Indonesia, Philippines and Thailand between 9 and 10; Malaysia 6; Singapore 5.
  - Top 5 benchmark scoring countries: Czech Republic, Israel, New Zealand, Hungary and Sweden.
  - Maximum score of 15 (index dimensions described in source).
- Empirical relation:
  - Time-varying estimated forward-looking coefficient β_{1,t} and the contribution β_{1,t} π^*_t are positively correlated with each country’s DE transparency score.
  - Results support that greater transparency is associated with higher degree of forward-looking dynamics.

---

### Inflation dynamics and cyclical fluctuations (Phillips Curve slope)
- Slope of Phillips Curve (β_{2,t}) is key for link between economic slack and inflation.
- Median contribution of output gap to inflation:
  - Fluctuated significantly over the last three decades.
  - More important ahead of the GFC; very limited role thereafter except in recessionary episodes.
- Phillips curve flattening:
  - All ASEAN-5 countries experienced a flattening of the Phillips curve in recent years.
  - Timing and magnitude vary by country (e.g., Indonesia: decline from average historical levels; others: large pre-2008 decline followed by stabilization at a lower level).
  - ASEAN-5 flattening is consistent with evidence for many advanced economies post-GFC, despite ASEAN-5 being relatively less affected by the crisis.
  - Contrast with South Korea: experienced mild but steepening of the Phillips curve since 2010 (compensating earlier flattening).
- Policy implication highlighted: flattening helps explain relatively low inflation in ASEAN-5 in 2017 despite growth momentum.

---

### Inflation dynamics, non-oil-import and oil price inflation
- Combined contributions of non-oil-import and oil price inflation declined substantially after the AFC.
- Non-oil import inflation declined markedly since early 2000s; oil price inflation contribution rose after AFC.
- Historical ranges for oil price contributions (median contribution to headline inflation):
  - 1996–2001: oil price contributions lie between 0.30 percentage points and -0.49 percentage points.
  - 2002–08: contributions range increases to (1.41 -0.17 percentage points).
  - Recent disinflation: oil price inflation drove down median inflation by -0.46 percentage points in 2015 and -0.59 percentage points in 2016.
- Overall synthesis:
  - Long-run inflation expectations became the most important driver.
  - Output gap movements have more limited impact in normal conditions but stronger during crises.
  - Import inflation contribution has become more limited over time, and relatively stable since early 2000s.

---

### Country-specific evidence — Indonesia (illustrative highlights)
- Indonesian inflation:
  - Declined from an average of 8.5% before the GFC to around 5% after the GFC.
  - Towards the end of the sample, disinflationary pressures limited, possibly reflecting fuel and electricity subsidies (reduced since 2015).
- Change in drivers over time (examples preserving exact figures from source):
  - In 2002:
    - Indonesian inflation ≈ 12%.
    - Contributions: economic activity 3.5 percentage points; non-oil-import prices 2.1 percentage points; oil price inflation 0.8 percentage points; expectations 3.6 percentage points.
  - In 2015:
    - Headline inflation 6.4%.
    - Contributions: expectations 4.9 percentage points; economic slack 0.05 percentage points; non-oil-import inflation 0.8 percentage points; oil price inflation -0.5 percentage points.
- Time-varying coefficients:
  - Contribution of trend inflation expectations in Indonesia rose sharply in the first half of the 2000s and stabilized thereafter.
  - The contribution of forward-looking dynamics increased by 74% from 2001 (2.67 percentage points) to 2005 (4.7 percentage points) and continued to narrowly fluctuate thereafter.

_Italic source attribution: IMF Working Paper — wp18147 (INTRODUCTION) — content unit as provided_

### 4.7 percentage points since then. The reason for the increase is twofold. First, the coefficient

### 4.7 percentage points since then. The reason for the increase is twofold. First, the coefficient

### Indonesia: drivers of inflation dynamics and policy developments
- Increase in forward-looking component:
  - Forward-looking coefficient rose from 0.32 in 2001 to 0.73 in 2007, stabilizing at around 0.7.
- Trend inflation estimates:
  - Indonesian trend inflation fell from 7.9% in 2001 to 5% from 2007 onwards.
- Monetary policy framework and communication:
  - Improvements in BI’s framework and communication between 2002-07 associated with rise in forward-looking dynamics.
  - Central Bank Act of 2004 introduced accountability and transparency measures: announcement of the inflation target and a monetary policy plan at the start of each year, quarterly Monetary Policy Report to parliament, publication of Monetary Board meeting decisions, economic forecasting models and a monetary policy outlook.
  - In early 2005 BI began to use interest rates as the main policy instrument due to difficulties controlling base money.
  - Mid-2010: BI re-evaluated its framework, adopting a policy mix of monetary and macroprudential policies; communication shifted to focus more on the medium-term inflation target and de-emphasizing other policy objectives.
- Headline inflation and contributors since the GFC:
  - Headline inflation stable around 5% since the GFC.
  - Real economic slack role limited since the GFC:
    - Output gap contributions declined from a range of 4.2 to -1.7 percentage points in 2001-07 to 0.4 to -0.7 percentage points in 2008-16.
    - Phillips Curve slope declined from 0.49 in 2001 to 0.03 in 2010.
    - Range of Indonesian output gap declined from 4.2% to -1.8% between 2001-08 to 0.4% to 0.7% thereafter.
  - Import price inflation importance diminished:
    - Coefficient of import prices increased from 0.13 in 2001 to 0.31 in 2006, then stabilizes at around 0.08 from 2008-16.
    - Contributions of non-oil-import and oil price inflation declined from averages of 2.4 and 1.5 percentage points between 2001-08 to 0.83 and 0.13 percentage points thereafter.
  - Policy coordination:
    - Better coordination between central bank and government policy on administered prices after 2008 helped attenuate oil price pass-through.

### Malaysia: inflation stability and forward-looking dynamics
- Average headline inflation and decomposition:
  - Over the sample, Malaysian headline inflation averaged 2.5%, of which 2.08 percentage points are explained by forward-looking inflation expectations and 0.2 percentage points are related to economic slack and import inflation (output gap 0.02 percentage points, non-oil-import price inflation 0.06 percentage points, oil price inflation 0.11 percentage points).
- Role of monetary policy framework and communication:
  - Since mid-2003 monetary policy statements released quarterly; from 2006 statements released directly after monetary policy meetings.
  - Central Bank Act in 2009 redefined and expanded BNM framework; post-2010 MPC meetings occur every two months.
- Changes in forward-looking dynamics:
  - Contribution of inflation expectations increased from an average of 1.8 percentage points between 2002-07 to an average of 2.3 percentage points over 2008-16.
  - Trend inflation estimates stable around 2.5% throughout sample.
  - Forward-looking coefficient rose from 0.4 over 2002-05 to 0.8 over 2006-16.
- Economic slack and import price contributions:
  - Economic slack impact limited except during GFC:
    - Decline in output during GFC contributed between 3.2 to -1.6 percentage points over 2008-10.
    - Slope parameter roughly 0.4 across sample but increases to 0.8 over 2008-10.
  - Non-oil-import and oil prices limited impact:
    - Contributions of import inflation range from 0.46 to -0.24 percentage points for non-oil-import and 0.37 to -0.26 percentage points for oil price inflation over entire sample.
    - Coefficient of import inflation increases from 0.006 over 2002-08 to 0.1 over 2009-16.
- Potential future influences:
  - Gradual liberalization of energy prices and introduction of a Goods and Service Tax via the fiscal act in 2010 might influence future inflation volatility and pass-through.

### Philippines: rising role of forward-looking expectations and changing slack sensitivity
- Importance of forward-looking dynamics:
  - Forward-looking dynamics account for 72% of headline inflation on average (3.6 percentage points of 4.9% headline inflation) over the entire sample.
  - Over 1996-2007, 64% of average headline inflation (3.64 percentage points of 5.8% headline inflation) explained by forward-looking dynamics.
  - Since GFC, share increased to 87% (3.4 percentage points of 3.9% headline inflation over 2008-16).
- Trend inflation and long-run expectations:
  - Long-run trend inflation declined from around 4% between 1995-2008 to 2.9% in 2016.
  - Coefficient on trend expectations stable around 0.4 across entire sample.
  - Long-run trend expectations were somewhat stable but below BSP official inflation target of 3% over 2014-16.
- Economic slack sensitivity and volatility:
  - Coefficient of economic slack increased from 0.03 in 1996 to 0.81 in 2008, indicating steepening of the Phillips Curve.
  - Reduced volatility of the output gap has offset increased sensitivity:
    - Standard deviation of output gap declined from 1.33 over 1996-2009 to 0.74 from 2010 onwards.
    - Contribution of economic slack over entire sample ranges between 2.71 to -1.29 percentage points.
- Import price contributions:
  - Contribution of import prices to headline inflation declined markedly:
    - Over 1996-2002, on average 28% of headline inflation associated with non-oil-import inflation (1.7 percentage points of 6% average headline inflation).
    - Between 2003-16 relative contribution declined to 3% (0.1 percentage points of 4.4% average headline inflation).
  - Oil price contributions relatively stable, ranging from 1 to -0.6 percentage points.
  - Coefficient of import inflation rises from 0.1 in 1996 to 0.41 in 2006 and stabilizes around 0.23 thereafter.
- Policy implications and risks:
  - Continued inflation expectations below official target increases risk of de-anchoring.
  - Against expected lower oil prices, recurrent disinflation episodes may become more likely.

### Singapore: openness, volatility, and forward-looking mitigating effects
- Average inflation and decomposition:
  - Between 1996 and 2016, headline inflation averaged 1.6%, with forward-looking dynamics explaining 1.85 percentage points.
  - Contributions from economic slack, non-oil-import and oil price inflation are 0.16, 0.47 and -0.11 percentage points respectively.
- Volatility and supply shocks:
  - Headline inflation volatility increased after the GFC with strong disinflationary pressures from 2013 to 2016 driven by falls in non-oil-import and oil price inflation.
  - Inflation fell from 4.7% in 2012 to 1% in 2014, and into negative territory in 2015 and 2016.
- Forward-looking dynamics:
  - Contribution of inflation expectations increased from 0.2 percentage points in 1996 to an average of 2.2 percentage points between 2007 and 2016, explaining 88% of average headline inflation during that period.
  - Forward-looking coefficient rises from 0.19 in 1996 to 0.34 in 2016, with a drop to 0.05 in 2008 and rise to 0.39 in 2010.
  - Sudden movements in survey-based expectations noted (e.g., drop from 2.2% to 1.3% in 2004 and rise to 1.9% in 2009), suggesting limitations of survey-based consensus expectations.
- Role of cyclical conditions:
  - Average contribution of economic slack is 0.51 percentage points and modest except for the GFC.
  - Coefficient on real economic activity is 0.14 between 1996 and 2004, rises thereafter (text truncated at end of source excerpt).

*Source: IMF working paper content (excerpt).*

### 0.46 during the GFC and declines again to 0.05 in 2016.

### wp18147 - 0.46 during the GFC and declines again to 0.05 in 2016.

### Singapore: inflation composition and time-varying import pass-through
- Non-oil-import and oil price inflation are more important for Singapore’s headline inflation compared to other ASEAN-5 countries and increase over time.
- Average contribution of non-oil-import and oil price inflation:
  - 0.59 percentage points between 1996 and 2006
  - 1.9 percentage points from 2007 onwards
- Contributions of non-oil-imports price inflation are systematically higher than those of oil price inflation.
- Import-inflation coefficient evolution:
  - 0.06 in 1996
  - increases to 0.15 in 2014
  - reveals a decreasing trend thereafter
- Possible disinflationary factor: steady decline in property prices, reflected in a negative residual toward the end of the sample.

### Thailand: expectations, slack, import-price pass-through, and policy implications
- Structural change in forward-looking component:
  - Forward-looking dynamics explains more than half (53%) of headline inflation after the AFC versus 33% before the AFC.
  - Absolute contribution of forward-looking dynamics:
    - 2.4 percentage points from 2001 to 2010
    - decreases to 2 percentage points thereafter
  - Lower contribution reflects both lower coefficient estimates and a decline of long-run trend inflation.
- Headline inflation trend:
  - Declined from an average of 6.5% before the AFC to 2.5% thereafter.
- Episodes of deflation:
  - Thailand experienced deflation in 2008 and 2015 due to a persistently negative output gap from 2008 and increased pass-through of oil price decline from 2014 onwards.
- Economic slack (output gap) contributions:
  - Average contribution during AFC: -2.13 percentage points
  - Average contribution during GFC: -2.58 percentage points
  - Average contribution since 2010: -0.08 percentage points
  - Time-varying coefficient of economic slack:
    - 0.03 in 1995
    - increases to 0.54 in 2008
    - declines to 0.3 in 2015
- Import-price contributions and coefficients:
  - Absolute contributions (entire sample) of non-oil-import and oil price inflation are on average 0.54 and 0.75 percentage points, respectively.
  - Contribution of non-oil-import inflation declines from an average of 1.44 percentage points during the AFC to 0.47 percentage points for the remaining sample.
  - Coefficient of import inflation increases from 0.08 in 1996 to 0.2 in 2008 and remains stable thereafter.
- Policy framework and communication:
  - BOT operated under an inflation targeting regime since 2000, with ongoing development of transparency and communication.
  - Change from core target (0.5%-3% for quarterly average core inflation) to headline target (annual average of 2.5%±1.5%) in early 2015 coincided with intensifying disinflationary pressures from oil price declines.
  - Memorandum of Understanding requires the MPC to explain missing the inflation target annually and issue a letter outlining the period and policy response for returning inflation to the target band.
  - BOT’s expectations of inflation returning to target did not materialize in 2015 and 2016.
- Interpretation and recommendations:
  - Reduced importance of the forward-looking component could not offset the oil price shock that pushed Thailand into deflation.
  - Disinflationary pressures and deflation reflect both falling oil prices and an increase in inflation persistence.
  - Continuous communication and policy actions to manage expectations and better align them to the inflation target can help counteract adverse cost-push shocks.
  - Further guidance on envisaged actions to achieve the target over a given time horizon, critical assessment of past performance, detailed discussion of shocks, and elaboration of internal MPC debate (including dissenting views) can help guide private-sector expectations.
  - Announcing an inflation target over a longer horizon than one year may contribute to stronger anchoring of inflation expectations.

### Robustness checks: data choices and model specifications
- General result: qualitative findings are robust to a series of sensitivity analyses on the benchmark model (equation (1)).
- Different measures of macroeconomic indicators:
  - Output-gap measures:
    - Benchmark uses HP-filter; robustness check uses two-sided band-pass filter.
    - Differences are quantitatively marginal; qualitative results of coefficients and contributions are very similar.
    - Quantitative differences: coefficients of output gap slightly higher during the GFC; median contribution of time-variation of output gap and import price inflation marginally lower.
  - Inflation expectations measures:
    - Re-estimation using Consensus long-run inflation expectations yields qualitatively similar results.
    - Consensus survey expectations are systematically higher than long-run inflation trend estimates.
    - Median contributions: time-variation in parameters slightly larger for output gap and import price inflation, less pronounced for forward-looking component.
    - Estimated coefficients show little differences compared to benchmark.
    - Note: Singapore estimation results are exclusively based on long-term Consensus Forecasts due to unusual uncertainty in trend inflation estimation.
  - Import-price measures:
    - Re-estimation with import prices from the WEO database yields qualitative implications in line with benchmark.
    - Combined contribution of oil and import price inflation quantitatively similar, but weight shifts toward oil price inflation in that variation.
- Model specification checks:
  - Time-varying parameters vs constant parameters:
    - Ignoring time variation would miss significant variation in inflation contributions; constant parameters can both under- and overestimate contributions in different periods.
    - Time-variation of the Phillips Curve slope is most relevant during recessionary episodes.
    - Non-oil-import inflation tends to be overestimated under constant parameters.
  - Decomposing import-price inflation into non-oil-import and oil price inflation (four-parameter model):
    - Median contributions of residuals rise to around 30% versus roughly 10% in benchmark.
    - Contribution of oil price inflation is substantially altered; forward-looking dynamics contributions are quantitatively lower than benchmark.
    - Country-specific coefficients of non-oil-imports are systematically higher by roughly 0.05 but follow same dynamics.
    - For oil-exporting countries (Indonesia and Malaysia) coefficients of oil price inflation are significantly positive and show little time variation.
    - Significant oil price inflation coefficients range from 0.005 to 0.028 across countries and time.
    - Overall, fourth specification is qualitatively similar to benchmark.
  - Models including exchange rate:
    - Augmenting model with nominal exchange rate (local currency to US Dollar) shows limited pass-through to headline inflation.
    - Exchange-rate coefficient is not significant across countries and time.
    - Median and country-specific contributions of exchange rate are quantitatively very small.
    - Median and country-specific contributions of oil price inflation are larger than in the benchmark model.

*Source: wp18147 - 0.46 during the GFC and declines again to 0.05 in 2016.*

### CONCLUSIONS

### CONCLUSIONS

### Overview and methodology
- Focus: quantitative evidence on the evolution of consumer price inflation dynamics in ASEAN-5 economies since the AFC.
- Key inflation drivers analyzed: inflation expectations, economic slack, non-oil-import inflation, and oil price inflation.
- Method: estimate Phillips Curves for each country allowing for time-varying parameters.

### Major findings on inflation dynamics
- Monetary policy frameworks have played a strong role in shaping inflation dynamics in the ASEAN-5 countries.
- Inflation has become increasingly forward-looking (less persistent).
- Inflation expectations have become the most important component of inflation in all countries.
- Inflation expectations have gradually become better anchored.
- The forward-looking coefficient (lower inflation persistence) and the overall contribution of inflation expectations display a positive relation with the evolution of ASEAN-5 central bankś transparency as measured by Dincer and Eichengreen (2014).
- The quantitative impact of cyclical fluctuations has gradually diminished over time.
- Results point to a significant flattening of the Phillips curve in ASEAN economies since the GFC.
- Increased exchange rate flexibility and further economic integration may have reduced the effect of non-oil and oil import inflation on headline inflation.
- Singapore is an exception with a significant coefficient of -0.8. The coefficient does not change significantly over time.

### Country-specific observations and challenges
- The paper provides a more detailed analysis of individual country experiences as references for other Emerging Economies.
- Despite improvements in monetary policy frameworks making ASEAN-5 countries more resilient to price shocks, challenges remain:
  - Strong disinflation and in some cases outright deflation in oil-importing economies over 2014-16 indicate supply-side shocks have not been completely offset by the forward-looking component of the inflation process.
  - Thailand: the protracted period of low inflation following adverse price shocks partly reflects a weakening of the forward-looking dynamics and of the anchoring of long-term inflation expectations.
- The flattening of the Phillips curve helps explain the relatively low inflation experienced across ASEAN-5 economies in 2017 despite the significant acceleration of economic activity in all the countries in the region and worldwide.

*Source: CONCLUSIONS, wp18147 - CONCLUSIONS*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2018/wp18147.pdf_