## _wp14191

## Source details

**Canonical URL:** [_wp14191](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14191.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14191.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14191.pdf.json)

---

### I. INTRODUCTION
- Inflation has declined sharply across Europe since end-2011.
- The 12-month inflation is well below the ECB’s price stability objective in the euro area (EA) and its target in the Czech Republic, Hungary, Poland, and Sweden and to a lesser extent in Romania.
- Several non-euro area EU countries experienced deflation:
  - Bulgaria: prices falling at an average 12-month rate of 1.3 percent since mid-2013.
  - Bosnia and Herzegovina: prices falling at an average 12-month rate of 1.2 percent since mid-2013.
  - Montenegro: slipped into milder deflation in 2014.
- Research question: whether non-EA EU countries are affected by undershooting of the EA inflation target or by other common factors.
- Approach: analyze inflationary developments in non-euro area EU countries within an open-economy, New Keynesian Phillips curve framework, controlling for imported inflation and focusing on the link between domestic inflationary pressures and those in the euro area.
- Paper structure overview:
  - Section II: model underpinning the analysis.
  - Section III: dynamics of key demand and supply-side disinflationary drivers.
  - Section IV: regression analysis results, variance decomposition of inflation, and 12-month-ahead inflation projections.
  - Section V: summary and policy implications.

### II. ANALYTICAL FRAMEWORK
- Model specification:
  - Open-economy New Keynesian Phillips curve (Galí and Gertler, 1999) with forward-looking inflation, inertia, demand and supply-side shocks.
  - Imported inflation decomposed into: nominal effective exchange rate, world food and oil prices, core inflation in the euro area.
- Key constructs and variables:
  - Headline inflation (HICP), expectation of future inflation (Consensus Economics two-year ahead mean), unemployment gap (Baxter-King bandpass filtered cyclical component), country-specific supply-side shocks, common external supply-side shocks (including imported inflation).
- Country groups:
  - CEE exchange-rate targeting (CEE-ET): Bulgaria, Croatia, Lithuania (peg currencies to the euro).
  - CEE inflation-targeting (CEE-IT): Czech Republic, Hungary, Poland, Romania.
  - EU3: Denmark, Sweden, United Kingdom.

### III. DATA OVERVIEW — DEMAND, SUPPLY, TRADE
- Demand-side findings:
  - Low inflation coincided with elevated unemployment.
  - Unemployment rates in deflation-experiencing countries (mostly exchange rate peggers) are trending up or near post-crisis peaks.
  - Labor market dynamics more favorable in CEE-IT and across Denmark, Sweden, and the United Kingdom.
  - Output gaps remain negative across EU3, CEE-ET, and CEE-IT groups; 2013 pick-up in domestic demand was insufficient to offset external deflationary pressures.
- Supply-side / cost-push findings:
  - Falling world food and energy prices and related cuts in administered energy prices have dampened inflation in all non-euro area EU countries since end-2011.
  - Share of food and energy in consumer baskets:
    - CEE: 44 percent
    - More developed peers: 25 percent
  - Better-than-average harvests in Romania, Hungary, Bulgaria eased food price inflation.
  - Since September 2012, core inflation in the euro area and non-euro area EU countries has decoupled from the average in the rest of the world; core inflation in CEE euro peggers has even turned negative.
  - Euro area is a major trade partner:
    - Accounts for about half of exports and imports of goods for CEE.
    - Accounts for about one third of exports and imports of goods for the rest of the non-euro area region.
  - Energy intensity: average energy intensity in CEE is close to double that of the euro area.
- Mechanical pass-through (upper-bound) accounting exercise (assumes one-to-one pass-through and share of euro-area sourced goods in consumer baskets similar to share of euro area value added in domestic demand):
  - Since end-2011, domestic-currency equivalent of the unit value of consumer goods imported from the euro area declined cumulatively by:
    - 6 percentage points in CEE inflation targeters
    - 4½ percentage points across CEE euro peggers
    - 3 percentage points in the rest of the region
  - Implication: since end-2011, lower EA import prices could have accounted for:
    - almost 20 percent of the decline in headline inflation across CEE
    - almost 10 percent in the rest of the region
  - Note: upper-bound given typical findings of low pass-through rates in empirical literature (Burstein and Gopinath, 2014).
- Trade openness and exchange-rate regime effects:
  - Share of foreign value-added content in aggregate demand:
    - Between 30 and 40 percent across CEE (Poland and Romania near the lower bound).
    - Much lower in other non-euro area EU countries.
  - Smaller CEE countries are more exposed to disinflationary spillovers from the euro area.
  - Exchange rate rigidity increases susceptibility to inflationary spillovers; regression slope linking tradable goods inflation in the euro area and CEE exchange targeters is steeper than for CEE-IT peers.

### IV. EXCHANGE-RATE DEVELOPMENTS
- NEER developments:
  - Since ECB President Draghi’s “whatever it takes” speech in July 2012, the euro nominal effective exchange rate (NEER) has appreciated by 10 percent.
  - NEERs of exchange-rate targeting CEE countries and Denmark appreciated, on average, at roughly half that rate.
  - The Polish złoty, Hungarian forint and Swedish krona initially staged stronger rallies than CEE peers, but have since partially retreated.
  - Easing cycles of monetary policies have added depreciating pressures on flexible exchange rates, partially offsetting deflationary external pressures.
  - NEER developments affect the domestic equivalent of world commodity prices and prices of non-energy industrial goods and services; impact likely larger across CEE because share of foreign value-added embodied in domestic demand is more than twice higher than in the euro area.
  - Footnote: In the euro area, a permanent 10 percent NEER depreciation is estimated to have a cumulative upward impact on non-energy industrial goods and services inflation of, respectively 0.6 and 0.5 percentage points over three years.

### V. EMPIRICAL STRATEGY AND ESTIMATION
- Sample and data:
  - Quarterly data for ten non-euro area EU countries over 2004–14; sample 2004q1-2014q1; N (observations) = 390.
- Regression specification:
  - Expectations-augmented open-economy New Keynesian Phillips curve with lags of inflation, expected inflation (2-year ahead), unemployment gap, NEER, contribution of taxes and administered prices, world food and oil inflation interacted with weights in consumer baskets, euro area core inflation (stand-alone or interacted).
  - Differentiated euro area impact via country-specific coefficients, interactions with share of foreign value-added in domestic demand, and exchange-rate regime dummies.
- Key estimation notes:
  - Fixed-effects OLS preferred (Model 5 is preferred specification).
  - Robustness: System 2SLS and 3SLS estimates, Huber-White standard errors, alternative euro-area pressure proxies, exclusion of UK — main findings preserved.
  - Instruments (for System 2SLS/3SLS): third lags of euro area output gap, euro area core inflation, global core inflation outside EA, first two principal components of country realizations; 13 instruments per equation.

### VI. MAIN REGRESSION RESULTS (selected coefficients; sample 2004q1-2014q1)
- Persistence and expectations:
  - Inflation (-1): 0.98 (0.049)*** (Model 1); 0.85 (0.046)*** (Model 2); 0.81 (0.047)*** (Model 3); 0.83 (0.046)*** (Model 4); 0.82 (0.046)*** (Model 5).
  - Inflation (-2): -0.34 (0.043)*** (Model 1); -0.28 (0.040)*** (Model 2); -0.27 (0.040)*** (Model 3); -0.28 (0.040)*** (Model 4); -0.28 (0.040)*** (Model 5).
  - Inflation expectations (2-year ahead): 0.48 (0.070)*** (Model 1); 0.63 (0.083)*** (Model 2); 0.61 (0.084)*** (Model 3); 0.59 (0.082)*** (Model 4); 0.60 (0.082)*** (Model 5).
- Demand and administered prices:
  - Unemployment gap: -0.13 (0.045)*** (Model 1); -0.18 (0.039)*** (Model 2); -0.16 (0.041)*** (Model 3); -0.18 (0.039)*** (Model 4); -0.17 (0.039)*** (Model 5).
  - Contribution of administered prices and taxes: 0.26 (0.036)*** (Model 1); 0.34 (0.037)*** (Model 2); 0.35 (0.037)*** (Model 3); 0.36 (0.037)*** (Model 4); 0.36 (0.036)*** (Model 5).
- Exchange rate and global factors:
  - Nominal effective exchange rate: -0.04 (0.007)*** (Model 1); -0.05 (0.007)*** (Models 2–5).
  - World oil inflation * weight of energy: 0.06 (0.012)*** (Model 2); 0.07 (0.012)*** (Models 3–5).
  - World food inflation * weight of food: 0.03 (0.011)*** (Models 2–5).
- Euro area core inflation effects:
  - Stand-alone (Model 2): 0.56 (0.126)***.
  - Country interactions (Model 3), selected:
    - Bulgaria: 1.44 (0.360)***.
    - Croatia: 1.19 (0.329)***.
    - Czech Republic: 0.81 (0.332)**.
    - Hungary: 0.99 (0.335)***.
    - Lithuania: 1.27 (0.376)***.
    - Poland: 0.40 (0.333).
    - Romania: 0.56 (0.335)*.
    - Denmark: 0.14 (0.325).
    - Sweden: 0.00 (0.33).
    - UK: -0.09 (0.325).
  - FX-regime interactions (Model 4):
    - Pegged: 0.92 (0.192)***.
    - Other arrangement: 0.73 (0.199)***.
    - Free floating: 0.07 (0.197).
  - FX-regime * foreign value-added (Model 5):
    - Pegged * foreign value-added in demand: 2.95 (0.564)***.
    - Other arrangement * foreign value-added in demand: 2.18 (0.563)***.
    - Free floating * foreign value-added in demand: 0.46 (0.782).
- Goodness of fit and tests:
  - R2: 0.94 (Model 1); 0.91 (Models 2–5).
  - N (observations): 390 (all models).
  - Reported F-statistics across specifications: F(9, 371) = 1.94; F(9, 362) = 2.68; F(9, 369) = 2.64; F(9, 370) = 3.67; F(9, 336) = 1.43.
  - Time dummies joint significance: t = 1.76 (0.078)* (reported for one test).

### VII. VARIANCE DECOMPOSITION AND COUNTRY HETEROGENEITY
- Overall contributions (preferred model, Model 5):
  - World food and energy prices together with related changes in administered prices and taxes account for about half of the variability of headline inflation across non-euro area EU countries.
  - Imported price pressures from the euro area account for between 2 and 37 percent of inflation variability in non-euro area EU countries over the whole sample period; these shares remained roughly unchanged in the last two years of the sample.
- Country exposure ranking (most to least susceptible to EA spillovers):
  - Bulgaria, Lithuania, Croatia most susceptible; followed by Hungary and the Czech Republic.
  - Poland, Romania, Sweden, United Kingdom relatively small disinflationary spillovers over full sample.
- Determinants across groups:
  - Over whole sample:
    - Nominal effective exchange rate explains about a third of inflation variance in inflation-targeting countries; contribution is much smaller across euro peggers.
    - Unemployment gap accounts for an additional 15 percent of inflation variance across CEE; much smaller effect in other non-euro area EU countries.
  - Since end-2011:
    - Falling world food and energy prices and related cuts in administered prices and taxes account for more than half of the volatility of headline inflation across non-euro area EU countries.
  - Recent period (post-2012q1):
    - NEER has played a larger role than over the whole sample period due to currency appreciations after July 2012 and weakening of some emerging market currencies since May 2013.

### VIII. ROBUSTNESS AND ADDITIONAL ECONOMETRIC FINDINGS
- Robustness checks:
  - Excluding the United Kingdom has no material effect on coefficients.
  - Adding global core inflation outside the euro area is not statistically significant; euro area core inflation effects preserved.
  - Substituting euro area core inflation with euro area output gap, unemployment gap, or instrumented euro area inflation does not change findings.
  - Huber-White adjusted standard errors leave significance unchanged.
- System estimations:
  - System 2SLS and 3SLS estimates generally smaller in magnitude than fixed-effects OLS but remain statistically significant at 95 percent for all explanatory variables except world food inflation.
  - System 2SLS confirms ranking by exchange rate regime of elasticities; System 3SLS reverses relative standing of euro peggers and other managed arrangements.
  - In all estimators, difference in elasticity of domestic inflation with respect to euro area core inflation between euro peggers and other managed arrangements is not statistically significant at 90 percent confidence.
- Note on estimators: Arellano-Bond dynamic-panel GMM not appropriate for this database due to small cross-sectional dimension.

### IX. INFLATION OUTLOOK AND CONDITIONAL FORECASTS
- Forecast setup:
  - Conditional on current policy settings—no changes in NEERs and taxes, no contemporaneous adjustment in administered energy prices.
  - Uses WEO projections of commodity prices and GDP growth, Consensus Forecasts (April 2014), and EA core inflation projection from the ECB (2014b).
  - Annual inflation in mid-2015 forecasted; lagged dependent variable = average of 12-month inflation over June 2013–May 2014.
  - Unemployment forecasts derived via calibrated Okun’s Law: ∆ Unemployment rate = 1.0 - 0.7 x GDP growth.
- Key forecast findings:
  - Over the next 12 months inflation likely to rebound very gradually in most non-euro area EU countries.
  - Firming recovery, gradual closing of unemployment and output gaps, well anchored expectations, and euro area recovery expected to provide a small boost to inflation.
  - Under current policies, inflation expected to remain low and below target in the next 12 months in several non-euro area EU countries.
  - Point forecast of the 12-month ahead inflation is around one percent across non-euro area EU countries, except in Romania and the United Kingdom.
  - In Hungary, Poland, and Sweden, inflation appears likely to remain below its target at least until mid-2015.
  - Inflation target considered within reach if it falls within the 95-percent confidence interval of the forecast; national targets are based on national CPI definitions (HICP comparison illustrative only).

### X. POLICY IMPLICATIONS AND SUMMARY
- Main drivers:
  - Disinflationary trends are primarily driven by cost-reducing external factors: falling world food and energy prices, cuts in administered prices, and taxes account for more than half of headline inflation volatility since end-2011.
  - Disinflation spillovers from the euro area have been important for exchange-rate targeters and inflation-targeters with high foreign value-added in domestic demand.
- Exchange rate regime and exposure:
  - Exchange rate regime is not the only determinant of susceptibility to EA spillovers; many exchange rate targeters are small open economies with high foreign value-added shares.
  - Euro peggers are proportionately more affected by ECB monetary policy and euro area inflation because of the lack of exchange rate buffer and high share of foreign value-added in domestic demand.
- Monetary policy trade-offs:
  - Euro peggers lack monetary autonomy.
  - Inflation-targeting countries must weigh trade-offs: risk of a self-feeding loop between disinflationary expectations and falling inflation against capital flow and financial stability concerns.
  - Indicators suggest current risk of unanchoring of inflationary expectations remains low throughout the region, conditional on long-term expectations and EA inflation not drifting lower.
  - Capital flows matter for policy space where non-resident investor presence is significant; capital inflows may resume as risk appetite returns or EA yields fall; outflows may resume with geopolitical concerns or unexpected Fed tightening.
  - Financial stability concerns are typically addressed with macroprudential tools; central banks may need to “lean against the wind” where macroprudential policies are not effective.

*Source: IMF staff compilation from document titled "_wp14191 - References ............................................................................................................."*

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

### _wp14191 - References .............................................................................................................

### I. INTRODUCTION
- Inflation has declined sharply across Europe since end-2011.
- At present, the 12-month inflation is well below the ECB’s price stability objective in the euro area (EA) and its target in the Czech Republic, Hungary, Poland, and Sweden and to a lesser extent in Romania.
- In several European countries, the region-wide spell of low inflation has turned into an outright deflation:
  - Bulgaria: prices have been falling at an average 12-month rate of 1.3 percent since mid-2013.
  - Bosnia and Herzegovina: prices have been falling at an average 12-month rate of 1.2 percent since mid-2013.
  - Montenegro slipped into a milder deflation in 2014.
- Research question: whether non-EA EU countries are affected by undershooting of the EA inflation target or by other common factors.
- Approach: analyze inflationary developments in non-euro area EU countries within an open-economy, New Keynesian Phillips curve framework, controlling for imported inflation and focusing on the link between domestic inflationary pressures and those in the euro area.
- Paper structure:
  - Section II: model underpinning the analysis.
  - Section III: dynamics of key demand and supply-side disinflationary drivers.
  - Section IV: regression analysis results, variance decomposition of inflation, and 12-month-ahead inflation projections.
  - Section V: summary and policy implications.

### II. ANALYTICAL FRAMEWORK
- Model: open-economy New Keynesian Phillips curve (Galí and Gertler, 1999) with forward-looking inflation, inertia, and demand and supply-side shocks.
- Key variables and constructs:
  - inflation (headline)
  - expectation of future inflation
  - unemployment gap as a measure of demand-side shocks (cyclical component of the unemployment rate extracted with the Baxter-King bandpass filter; expect δ < 0)
  - vector of country-specific supply-side shocks
  - vector of common external supply-side shocks, including imported inflation
- Imported inflation is decomposed into:
  - nominal effective exchange rate
  - world food and oil prices
  - core inflation in the euro area
- Country grouping for analysis:
  - CEE exchange-rate targeting (CEE-ET): Bulgaria, Croatia, Lithuania (all peg currencies to the euro)
  - CEE inflation-targeting (CEE-IT): Czech Republic, Hungary, Poland, Romania
  - EU3: Denmark, Sweden, United Kingdom
- Accounts from national central banks and the ECB indicate energy, food, and administered prices as key drivers; also highlighted are falling import prices, weakness of domestic demand, and exchange rate appreciation in specific countries.

### III. OVERVIEW OF THE DATA
#### A. Demand-Side Factors
- Low inflation coincided with elevated unemployment.
- Unemployment rates of countries experiencing deflation—most of which are exchange rate peggers—are either trending up or are near their post-crisis peaks.
- Labor market dynamics are more favorable in CEE inflation-targeting countries and across Denmark, Sweden, and the United Kingdom.
- The 2013 gradual pick-up of domestic demand was insufficient to offset deflationary external pressures; output gaps remain negative across EU3, CEE-ET, and CEE-IT groups.

#### B. Supply-Side, Cost-Push Factors
- Falling world food and energy prices and related cuts in administered energy prices have dampened inflation in all non-euro area EU countries since end-2011.
- Share of food and energy in consumer baskets:
  - CEE: 44 percent
  - More developed peers: 25 percent
- Better-than-average harvests (e.g., Romania, Hungary, Bulgaria) further eased food price inflation.
- EU-wide decline in core inflation suggests possible disinflationary spillovers from the euro area; since September 2012, core inflation in the euro area and non-euro area EU countries has decoupled from the average in the rest of the world.
- In CEE euro peggers, core inflation has even turned negative.
- Euro area as major trade partner:
  - Accounts for about half of exports and imports of goods for CEE.
  - Accounts for about one third of exports and imports of goods for the rest of the non-euro area region.
- Energy intensity: average energy intensity in CEE is close to double that of the euro area, amplifying effects of energy-price-driven cost changes.
- Mechanical pass-through accounting exercise (upper-bound estimate, assuming one-to-one pass-through and share of euro-area sourced goods in consumer baskets similar to share of euro area value added in domestic demand):
  - Since end-2011, the domestic-currency equivalent of the unit value of consumer goods imported from the euro area declined cumulatively by:
    - 6 percentage points in CEE inflation targeters
    - 4½ percentage points across CEE euro peggers
    - 3 percentage points in the rest of the region
  - Implication: since end-2011, lower EA import prices could have accounted for:
    - almost 20 percent of the decline in headline inflation across CEE
    - almost 10 percent in the rest of the region
  - Note: presented as an upper-bound given typical findings of low pass-through rates in empirical literature (Burstein and Gopinath, 2014).

#### C. Trade Openness and Exchange Rate Regime Effects
- Share of foreign value-added content in aggregate demand:
  - Between 30 and 40 percent across CEE (countries with larger internal markets, such as Poland and Romania, near the lower bound).
  - Much lower in other non-euro area EU countries.
- Implication: smaller CEE countries are more exposed to disinflationary spillovers from the euro area through the trade channel.
- Exchange rate rigidity increases susceptibility to inflationary spillovers from trade partners.
- Empirical observation: slope of regression line linking tradable goods inflation in the euro area and CEE exchange targeters is steeper than for CEE inflation-targeting peers.

### IV. EXCHANGE-RATE DEVELOPMENTS
- Appreciation of currencies pegged to the euro magnified pass-through of lower international prices to domestic inflation since mid-2012.
- (Figures and time-series evidence referenced: Figure 5 and associated plots of unit value of imports from euro area in domestic currency by country groups.)

### V. SUMMARY OF ANALYTICAL APPROACH AND EMPIRICAL STEPS
- Model specification follows Galí and Gertler (1999) augmented for imported inflation (as in Galí and Monacelli, 2005) and aligned with IMF (2013a) analysis.
- Empirical strategy includes:
  - Controlling for unemployment gap (Baxter-King bandpass filtered cyclical component).
  - Controlling for imported inflation via nominal effective exchange rates, world food and oil prices, and euro area core inflation.
  - Group-specific analysis by monetary policy regime (CEE-ET vs CEE-IT) and separate treatment of Denmark, Sweden, United Kingdom.
  - Regression analysis to decompose inflation variance into domestic and external contributions and produce 12-month-ahead projections.

*Source: IMF staff compilation from the document titled "_wp14191 - References ............................................................................................................."*

### Appendix Figure 2). Since ECB President Draghi’s “whatever it takes” speech in July 2012,

### _wp14191 - Appendix Figure 2). Since ECB President Draghi’s “whatever it takes” speech in July 2012,

### Nominal Effective Exchange Rate (NEER) developments
- Since ECB President Draghi’s “whatever it takes” speech in July 2012, the euro nominal effective exchange rate (NEER) has appreciated by 10 percent.
- The NEERs of exchange-rate targeting CEE countries and Denmark also appreciated, on average, at roughly half that rate.
- The Polish złoty, Hungarian forint and Swedish krona initially staged stronger rallies than their CEE peers, but have since partially retreated.
- Over time, the easing cycle of monetary policies has added to the depreciating pressures on flexible exchange rates in many countries in the region, partially offsetting the deflationary external pressures.
- NEER developments affect:
  - the domestic equivalent of world commodity prices;
  - prices of non-energy industrial goods and services.
- The impact is likely larger across CEE because the share of foreign value-added embodied in their domestic demand is more than twice higher than in the euro area.
- Footnote finding: In the euro area, a permanent 10 percent NEER depreciation is estimated to have a cumulative upward impact on non-energy industrial goods and services inflation of, respectively 0.6 and 0.5 percentage points over three years.

### Portfolio re-allocations, capital flows and global drivers
- Portfolio re-allocations by non-residents have been an important driver of exchange rates in inflation-targeting EU countries.
- Post-July 2012 appreciation of the euro and CEE currencies initially reflected capital inflows driven by improved investor sentiment as tail risks were re-evaluated.
- Since May 2013, strength of the euro and CEE currencies pegged to it in NEER terms has been supported by depreciation of currencies of some large emerging markets (Brazil, India, South Africa) and of inflation-targeting CEE countries and Sweden.
- Post-May 2013 depreciation of emerging market currencies occurred alongside reversal of capital inflows triggered by announcement of a gradual normalization of US monetary policy.
- Regression analysis (April 2014 Regional Economic Issues) shows benchmark interest rates in advanced economies and global investors’ risk aversion (proxied by the Chicago Board Options Exchange Market Volatility Index) play an important role in dampening/reversing capital inflows and pushing up government bond yields across CESEE.
- Empirical implication: monetary policies of advanced economies have a sizable impact on capital inflows and interest rates in CESEE.

### Effect of tax and administered price changes
- Contributions to inflation of indirect taxes and, where applicable, administered prices have been significant across non-euro area EU countries.
- Since end-2011, changes in indirect taxes and administered prices (mostly of energy) have contributed to easing of domestic price pressures in most countries.
- Energy price inflation has eased further in CEE due to dissipating base effects of previous hikes and follow-up cuts in administered prices of electricity and/or gas in Bulgaria, Croatia, the Czech Republic, Hungary, Latvia, Lithuania, and Poland.
- Many countries with steepest declines in headline inflation have lowered administered energy prices by a wider margin than peers.
- Changes in indirect taxes and administered prices have been less important for inflation developments in Denmark and Sweden.
- Country note: In Hungary, inflation dynamics are largely determined by the effects of government measures, with sharp deceleration of headline inflation recently explained in part by several cuts in regulated energy prices for households.

### Inflationary expectations
- Available survey data suggest, on balance, the risk of unanchoring of inflationary expectations currently remains low throughout the region.
- Differences exist between fixed-exchange rate and inflation-targeting regimes in formation of inflationary expectations: euro peggers import ECB monetary stance and credibility, making expectations more exogenous to domestic policies.
- Consensus forecasters (March–April 2014) put inflation two-year ahead:
  - close to target in the Czech Republic, Hungary, Poland, and Sweden;
  - above target in Romania and the United Kingdom.
- Expectations of inflation two-year ahead are also above the ECB target of 2 percent in CEE exchange-rate targeting countries.
- European Commission household surveys show households in CEE—including countries experiencing deflation—continue to believe by a wide margin that prices will rise over the next 12 months.
- Notable downward drift in medium-term inflationary expectations in a number of non-euro area EU countries is yet to be reversed.
- Sweden: July and August EC consumer surveys show more Swedish households now expect prices to fall over the next twelve months than to increase.
- Caveat: survey-based indicators can be wrong; long-term expectations were positive on the eve of three deflationary episodes in Japan (Moghadam, Teja, and Berkmen (2014)).

### Empirical analysis — regression specification and strategy
- Estimated model: open-economy New Keynesian Phillips curve with quarterly data for ten non-euro area EU countries over 2004–14.
- Regression includes sufficient lags of inflation to address inertia and residual autocorrelation.
- Regression variables include: expected inflation, unemployment gap, exchange rate appreciation/depreciation (NEER), contribution of taxes and administered prices to headline inflation, time effects, world commodity prices, and euro area core inflation.
- Differentiated impact of euro area price pressures allowed via:
  - country-specific coefficients;
  - interactions with share of foreign value-added in domestic demand;
  - exchange-rate regime dummy variables.

### Main regression findings (Table 1, selected results and exact coefficients)
- Key estimated coefficients (standard errors in parentheses; significance: * p<0.1; ** p<0.05; *** p<0.01). Sample 2004q1-2014q1.
- Country-specific factors:
  - Inflation (-1): 0.98 (0.049)*** (Model 1); 0.85 (0.046)*** (Model 2); 0.81 (0.047)*** (Model 3); 0.83 (0.046)*** (Model 4); 0.82 (0.046)*** (Model 5).
  - Inflation (-2): -0.34 (0.043)*** (Model 1); -0.28 (0.040)*** (Model 2); -0.27 (0.040)*** (Model 3); -0.28 (0.040)*** (Model 4); -0.28 (0.040)*** (Model 5).
  - Inflation expectations (2-year ahead): 0.48 (0.070)*** (Model 1); 0.63 (0.083)*** (Model 2); 0.61 (0.084)*** (Model 3); 0.59 (0.082)*** (Model 4); 0.60 (0.082)*** (Model 5).
  - Unemployment gap: -0.13 (0.045)*** (Model 1); -0.18 (0.039)*** (Model 2); -0.16 (0.041)*** (Model 3); -0.18 (0.039)*** (Model 4); -0.17 (0.039)*** (Model 5).
  - Contribution of administered prices and taxes: 0.26 (0.036)*** (Model 1); 0.34 (0.037)*** (Model 2); 0.35 (0.037)*** (Model 3); 0.36 (0.037)*** (Model 4); 0.36 (0.036)*** (Model 5).
  - Nominal effective exchange rate: -0.04 (0.007)*** (Model 1); -0.05 (0.007)*** (Models 2–5).
- Global factors:
  - World oil inflation * weight of energy in consumer baskets: 0.06 (0.012)*** (Model 2); 0.07 (0.012)*** (Models 3–5).
  - World food inflation * weight of food in consumer baskets: 0.03 (0.011)*** (Models 2–5).
- Euro Area core inflation (stand-alone, Model 2): 0.56 (0.126)***.
- Euro Area core inflation interacted with country dummies (Model 3), selected estimates:
  - Bulgaria: 1.44 (0.360)***.
  - Croatia: 1.19 (0.329)***.
  - Czech Republic: 0.81 (0.332)**.
  - Hungary: 0.99 (0.335)***.
  - Lithuania: 1.27 (0.376)***.
  - Poland: 0.40 (0.333).
  - Romania: 0.56 (0.335)*.
  - Denmark: 0.14 (0.325).
  - Sweden: 0.00 (0.33).
  - UK: -0.09 (0.325).
- Euro Area core inflation interacted with FX-regime dummy variables (Model 4):
  - Pegged: 0.92 (0.192)***.
  - Other arrangement: 0.73 (0.199)***.
  - Free floating: 0.07 (0.197).
- Euro Area core inflation interacted with FX-regime dummies and share of foreign value-added in demand (Model 5):
  - Pegged * foreign value-added in demand: 2.95 (0.564)***.
  - Other arrangement * foreign value-added in demand: 2.18 (0.563)***.
  - Free floating * foreign value-added in demand: 0.46 (0.782).
- Model fit and tests (exact figures reported):
  - R2 (Models 1–5): 0.94 (Model 1); 0.91 (Models 2–5).
  - N (observations): 390 (all models).
  - Time dummies joint significance: t = 1.76 (0.078)* reported for one test.
  - F-statistics reported: F(9, 371) = 1.94; F(9, 362) = 2.68; F(9, 369) = 2.64; F(9, 370) = 3.67; F(9, 336) = 1.43 (across model specifications).

### Interpretation of empirical results and variance decomposition
- Food and energy prices (world food and energy price changes) together with related changes in administered prices and taxes account for about half of the variability of headline inflation across non-euro area EU countries (variance decomposition based on preferred model, Table 1, Model 5).
- Disinflation spillovers from the euro area:
  - Imported price pressures from the euro area account for between 2 and 37 percent of inflation variability in non-euro area EU countries over the whole sample period.
  - These shares have remained roughly unchanged in the last two years of the sample.
- Country heterogeneity:
  - Bulgaria, Lithuania, and Croatia are most susceptible to price spillovers from the euro area, followed by Hungary and the Czech Republic—three of these have currencies pegged to the euro and all have high foreign value-added in domestic demand (Croatia data caveat noted).
  - Countries with more rigid exchange rate arrangements tend to import more inflation from the euro area.
  - Inflation spillovers are larger the higher the share of domestically consumed foreign value-added (e.g., Czech Republic, Hungary).
  - Disinflationary spillovers from the euro area to Poland, Romania, Sweden, and the United Kingdom are relatively small over the full sample, due to smaller foreign value-added in domestic demand and greater exchange rate flexibility of the Polish złoty, Swedish krona, and British pound.

### Preferred specification and robustness
- Model 5 (interaction of euro area core inflation with exchange-rate regime dummies and share of foreign value-added in demand) is the preferred regression specification.
- Model 5 achieves parsimony while matching R2 of the model with unrestricted country-specific euro area core inflation coefficients.

_Italic: Sources: Eurostat; EPFR data; Consensus Economics Long-Term Forecasts; April 2014 Regional Economic Issues (IMF, 2014a); ECB Monthly Bulletin, March 2014. Notes and sample: sample 2004q1-2014q1; s.e. in parentheses; * p<0.1; ** p<0.05; *** p<0.01._

### Appendix Figure 3). At the same time, the contribution of euro area core inflation to

### _wp14191 - Appendix Figure 3). At the same time, the contribution of euro area core inflation to

### Euro-area core inflation spillovers and country exposure
- The contribution of euro area core inflation to the variance of domestic inflation in these countries has increased since end-2011, particularly in Poland (Appendix Figure 3).
- In line with the findings in IMF (2014b), the effect of euro area price pressures on Swedish inflation is positive but small.
- For a given degree of exposure to foreign prices, the difference in the elasticity of domestic inflation with respect to euro area core inflation in euro peggers and countries with other managed arrangements is not statistically significantly at the 90 percent confidence level.
- Regression-based results (Table 1, Model 5; Appendix Table 3) indicate disinflationary spillovers from the euro area.

### Determinants of inflation variance across country groups
- Over the whole sample period:
  - The nominal effective exchange rate explains about a third of the inflation variance in inflation-targeting countries.
  - The nominal effective exchange rate contribution is much smaller across euro peggers.
  - The unemployment gap accounts for an additional 15 percent of inflation variance across CEE.
  - The unemployment gap effect is much smaller in the other non-euro area EU countries.
- Since end-2011:
  - Falling world food and energy prices and related cuts in administered prices, as well as taxes, account for more than half of the volatility of headline inflation across non-euro area EU countries.
- Recent developments:
  - In the last two years, the nominal effective exchange rate has played a larger role than over the whole sample period, reflecting the appreciation of many regional currencies in the aftermath of ECB President Draghi’s “whatever it takes” speech in July 2012 and the weakness of the currencies of some large emerging markets since May 2013.
- The limited impact of cyclical unemployment is noted and described as consistent with the tendency of flattening of the Phillips-curve across advanced economies in IMF(2013a).

### Robustness tests and econometric methods
- Excluding the United Kingdom from the sample has no material effect on regression coefficients (Appendix Table 2, Model 1).
- Global core inflation outside the euro area is not statistically significant, when added to the preferred regression specification (Appendix Table 2, Model 2). At the same time, the effects of euro area core inflation are preserved.
- Substituting euro area core inflation sequentially with the euro area output gap, unemployment rate gap, and instrumented euro area inflation (Appendix Table 2, Models 3 to 5) does not change the findings.
- Adjusting standard errors using the Huber-White sandwich estimator leaves statistical significance of the coefficients unchanged.
- To address endogeneity, Model 5 in Table 1 is estimated by System 2SLS and 3SLS (Appendix Table 3).
  - Instruments used (lagged by three periods): third lags of euro area output gap, euro area core inflation, global core inflation outside the euro area, and the first two principal components of the country-realizations in the sample of domestic inflation, unemployment gap, output gap, exchange rate appreciation/depreciation, and the contribution of taxes and administered prices to headline inflation. This gives 13 instruments for use on each equation in the system.
  - System 2SLS and 3SLS estimates of all regression coefficients, except for the lags of the dependent variable, are smaller in magnitude than their fixed-effects OLS counterparts but remain statistically significant at the 95 percent level of confidence for all explanatory variables, except for world food inflation.
  - System 2SLS confirms the relative ranking by exchange rate regime of the elasticities of domestic inflation with respect to euro area core inflation derived from fixed-effects OLS.
  - The relative standing of euro peggers and countries with other managed arrangements is reversed in the System 3SLS results.
  - In both System 2SLS and System 3SLS, and in line with fixed-effects OLS findings, the difference in the elasticity of domestic inflation with respect to euro area core inflation in euro peggers and countries with other managed arrangements is not statistically significant at the 90 percent confidence level.
- Note: The more commonly used Arellano-Bond dynamic-panel GMM estimators are not appropriate for this database because the cross-sectional unit dimension must be very large to rely on their asymptotic properties.

### Inflation outlook and conditional forecasts
- Forecast setup:
  - Forecasts constructed under current policy settings—assuming no changes in nominal effective exchange rates and taxes, and no contemporaneous adjustment in administered prices of energy given the small projected decline in commodity prices.
  - Uses WEO projections of commodity prices and GDP growth, inflation forecasts from Consensus Forecasts (as of April 2014), and EA core inflation projection from the ECB (2014b).
  - Annual inflation in mid-2015 is forecast using the model; the lagged dependent variable is calculated as the average of 12-month inflation rates over the period June 2013–May 2014.
  - Desks’ real GDP projections are converted into unemployment rate forecasts using a calibrated Okun’s Law type relationship: ∆ Unemployment rate = 1.0 - 0.7 x GDP growth.
- Key forecast findings:
  - Forecasts suggest that over the next 12 months inflation is likely to rebound, but very gradually in most non-euro area EU countries.
  - Expected firming of the economic recovery and gradual closing of the unemployment and output gaps, together with well anchored inflation expectations and euro area recovery, will provide a small boost to inflation.
  - Under current policies, inflation is expected to remain low and below target in the next 12 months in several non-euro area EU countries.
  - The point forecast of the 12-month ahead inflation is around one percent across non-euro area EU countries, except in Romania and the United Kingdom.
  - In Hungary, Poland, and Sweden, inflation appears likely to remain below its target at least until mid-2015.
  - The inflation target is considered to be within reach if it falls within the 95-percent confidence interval of the forecast. National inflation targets are set in reference to the national definition of the consumer price index; comparison between HICP forecast and national targets is illustrative only.

### Summary and policy implications
- Main drivers:
  - Disinflationary trends in non-euro area EU countries are primarily driven by cost-reducing, external factors.
  - Falling world food and energy prices, related cuts in administered prices, and taxes account for more than half of headline inflation volatility since end-2011.
  - Disinflation spillovers from the euro area have been an important factor for exchange-rate targeters and inflation targeters with high foreign value added in domestic demand.
- Exchange rate regime and exposure:
  - The exchange rate regime is not the main factor determining countries’ susceptibility to inflation spillovers from the euro area, as most exchange rate targeters are also small open economies with a high share of foreign value-added in domestic demand.
  - Euro peggers are proportionately more affected by ECB monetary policy and euro area inflation because of the lack of exchange rate buffer and high share of foreign value-added in domestic demand.
- Monetary policy trade-offs:
  - While euro peggers do not have monetary autonomy, the monetary policy response to low inflation in inflation-targeting countries needs to weigh the risk of a self-feeding loop between disinflationary expectations and falling inflation, capital flows, and financial stability concerns.
  - Judging by indicators of inflationary expectations, the risk of their unanchoring currently appears low throughout the region assuming long-term inflation expectations and actual inflation do not drift lower in the euro area. However, continued vigilance is warranted, as long-term inflation expectations were also positive on the eve of the three deflationary episodes in Japan.
  - In countries with significant presence of non-resident investors in local markets, capital flows have a bearing on the space for policy maneuver: capital inflows may pick up as risk appetite returns or as EA yields fall in response to ECB easing; there is also a risk that capital outflows may resume due to heightened geopolitical concerns or unexpected Fed tightening.
  - Financial stability concerns are typically best addressed with macroprudential tools, but central banks might have to “lean against the wind”, where such policies are not effective.

*Source: IMF staff calculations and analysis in the supplied content.*

### Appendix Figure 2. Nominal Effective Exchange Rate

### Appendix Figure 2. Nominal Effective Exchange Rate

### Nominal effective exchange rate and inflation dynamics
- Appendix Figure 2 presents the Nominal Effective Exchange Rate (Percent, year-over-year). (Figure label only; numerical series not reproduced in text.)
- Appendix Figure 3 and related panels decompose inflation variance for Non-EU EA Countries (2004-14), with contributions shown for:
  - Unemployment gap
  - Nominal effective exchange rate
  - Taxes and administered prices
  - World food and energy prices
  - Euro Area core inflation
- Recent period: (2012q1 - 2014q1, Contributions in percent) — country ordering appears in charts: Lithuania, Bulgaria, Czech Republic, Denmark, Romania, Poland, Hungary, Sweden, UK, Croatia.
- Full sample: (2004q1 - 2014q1) — country ordering appears in charts: Denmark, Bulgaria, Lithuania, Croatia, Czech Republic, Hungary, Sweden, UK, Poland, Romania.
- Appendix Figure 4 shows Consumer Inflationary Expectations over Next 12 Months, 2009-14 (Index), series shown for:
  - CEE-ET
  - CEE-IT
  - Central and Eastern Europe
  - Denmark, Sweden, United Kingdom
  - Index range in figures spans from -20 to 60 with ticks at -20, -10, 0, 10, 20, 30, 40, 50, 60 and dates Dec-08, Dec-10, Dec-12, Dec-14.

### Chronology of monetary policy measures (Non-Euro Area EU Countries, Jul 2012–2014)
- Appendix Table 1 provides a chronology of monetary policy measures in Non-Euro Area EU Countries, Jul 2012 - 2014, including:
  - Multiple policy rate cuts across 2012–2014 (examples preserved exactly as shown in timeline):
    - 2012 Jul 05: ▼25 bp to 0.20% ▲GBP 50 bln to 375 bln
    - 2012 Aug 28: ▼25 bp to 0.75%
    - 2012 Sep 06: ▼25 bp to 1.25%
    - 2012 Sep 27: ▼25 bp to 6.50%
    - 2012 Sep 27: ▼25 bp to 0.25%
    - 2012 Oct 30: ▼25 bp to 6.25%
    - 2012 Nov 01: ▼20 bp to 0.05%
    - 2012 Nov 07: ▼25 bp to 4.50%
    - 2012 Nov 27: ▼25 bp to 6.00%
    - 2012 Dec 05: ▼25 bp to 4.25%
    - 2012 Dec 18: ▼25 bp to 5.75% ▼25 bp to 1.00%
    - 2013 Jan 09: ▼25 bp to 4.00%
    - 2013 Jan 25: ▲10 bp to 0.30%
    - 2013 Jan 29: ▼25 bp to 5.50%
    - 2013 Feb 06: ▼25 bp to 3.75%
    - 2013 Feb 26: ▼25 bp to 5.25%
    - 2013 Mar 07: ▼50 bp to 3.25%
    - 2013 Mar 27: ▼25 bp to 5.00%
    - 2013 Apr 24: ▼25 bp to 4.75%
    - 2013 May 02: ▼25 bp to 0.50%
    - 2013 May 03: ▼10 bp to 0.20%
    - 2013 May 09: ▼25 bp to 3.00%
    - 2013 May 28: ▼25 bp to 4.50%
    - 2013 Jun 05: ▼25 bp to 2.75%
    - 2013 Jun 25: ▼25 bp to 4.25%
    - 2013 Jul 01: ▼25 bp to 5.00%
    - 2013 Jul 03: ▼25 bp to 2.50%
    - 2013 Jul 23: ▼25 bp to 4.00%
    - 2013 Aug 05: ▼50 bp to 4.50%
    - 2013 Aug 27: ▼20 bp to 3.80%
    - 2013 Sep 24: ▼20 bp to 3.60%
    - 2013 Sep 30: ▼25 bp to 4.25%
    - 2013 Oct 29: ▼20 bp to 3.40%
    - 2013 Nov 05: ▼25 bp to 4.00%
    - 2013 Nov 07: ▼25 bp to 0.25%
    - 2013 Nov 25: ▼125 bp to 5.00%
    - 2013 Nov 26: ▼20 bp to 3.20%
    - 2013 Dec 17: ▼25 bp to 0.75%
    - 2013 Dec 18: ▼20 bp to 3.00%
    - 2014 Jan 08: ▼25 bp to 3.75%
    - 2014 Jan 21: ▼15 bp to 2.85%
    - 2014 Feb 04: ▼25 bp to 3.50%
    - 2014 Feb 18: ▼15 bp to 2.70%
    - 2014 Mar 25: ▼10 bp to 2.60%
    - 2014 Apr 29: ▼10 bp to 2.50%
    - 2014 May 27: ▼10 bp to 2.40%
    - 2014 Jun 05: ▼10 bp to 0.15%
    - 2014 Jun 24: ▼10 bp to 2.30%
    - 2014 Jul 03: ▼50 bp to 0.25%
    - 2014 Jul 22: ▼20 bp to 2.10%
    - 2014 Aug 04: ▼25 bp to 3.25%
    - 2014 Sep 04: ▼10 bp to 0.05%
- Additional monetary policy measures (text excerpts preserved):
  - Euro area: The European central bank changed the deposit facility and marginal lending facility rates in lockstep with the changes made to the main refinancing operations fixed rate.
  - Croatia: On April 10, 2013, Hrvatska Narodna Banka lowered the interest rate on banks' overnight deposits from 0.25% to 0.00%.
  - On June 5, 2014, the ECB announced a prolongation of fixed rate, full allotment tender procedures, suspension of the weekly fine-tuning operation sterilising the liquidity injected under the Securities Markets Programme, and a decision to launch outright purchases of asset-backed securities in the coming months.
  - On August 1, 2013, the Bank of England introduced forward guidance, and announced its intention not to raise Bank Rate from its current level of 0.5% at least until the Labour Force Survey headline measure of the unemployment rate has fallen to a threshold of 7%, subject to three knockout conditions related to inflation forecast and expectations as well as financial stability.
  - On November 7, 2013, the CNB Board decided to start using the exchange rate as an additional instrument for easing the monetary conditions. The CNB intervenes on the foreign exchange market to weaken the koruna so that the exchange rate of the koruna against the euro is close to CZK 27.
  - On August 1, 2013, the Bank of England increased the size of its asset purchase program by GBP 50 billion to a total of GBP 375 billion in July 2012, and later extended it until January 2015 and further modified it in 2013.
  - The “Funding for Growth Scheme” was launched in June 2013 to improve SMEs access to bank financing by extending no-cost loans to banks for on-lending to SMEs at negative or close to zero real interest rates.
  - On April 25, 2014, the Danmarks Nationalbank increased its interest rate on certificates of deposit by 15 basis points to 0.05%.

### Econometric results: Expectations-augmented Phillips curve and robustness checks
- Appendix Table 2 — Fixed-Effects Estimation of Expectations-Augmented Phillips Curve with Alternative Measures of Euro Area Inflationary Pressures. Sample: 2004q1-2014q1. s.e. in parentheses; * p<0.1; ** p<0.05; *** p<0.01.
- Selected country-specific estimated coefficients (multiple specifications shown; parentheses contain standard errors):
  - Inflation (-1): 0.83 (0.049)*** ; repeated across columns as 0.83 (0.046)*** ; also 0.78 (0.050)***
  - Inflation (-2): -0.28 (0.042)*** ; -0.28 (0.039)*** ; -0.23 (0.041)*** ; -0.27 (0.040)*** ; -0.22 (0.042)***
  - Inflation expectations (2-year ahead): 0.60 (0.088)*** ; 0.60 (0.082)*** ; 0.46 (0.085)*** ; 0.52 (0.083)*** ; 0.53 (0.085)***
  - Unemployment gap: -0.18 (0.041)*** ; -0.17 (0.039)*** ; -0.19 (0.039)*** ; -0.17 (0.041)*** ; -0.23 (0.037)***
  - Contribution of administered prices and taxes: 0.36 (0.039)*** ; 0.35 (0.037)*** ; 0.39 (0.037)*** ; 0.37 (0.037)*** ; 0.41 (0.039)***
  - Nominal effective exchange rate: -0.05 (0.008)*** ; -0.05 (0.007)*** ; -0.05 (0.007)*** ; -0.05 (0.007)*** ; -0.04 (0.007)***
- Global factors (selected):
  - World oil inflation * weight of energy in consumer baskets: 0.07 (0.013)*** ; repeated 0.07 (0.013)*** ; 0.07 (0.012)*** ; 0.08 (0.012)*** ; 0.05 (0.013)***
  - World food inflation * weight of food in consumer baskets: 0.03 (0.011)*** ; 0.03 (0.011)*** ; 0.00 (0.012) ; 0.00 (0.013) ; 0.03 (0.011)**
  - World core inflation (excl. euro area): 0.14 (0.12) (not statistically highlighted)
- Proxies for Euro Area price pressures interacted with FX-regime dummy variables and the share of foreign value-added in demand (selected coefficients, s.e. in parentheses):
  - Euro Area core inflation:
    - Pegged * foreign value-added in demand: 2.90 (0.588)*** ; 3.08 (0.575)***
    - Other arrangement * foreign value-added in demand: 2.17 (0.586)*** ; 2.30 (0.57)***
    - Free floating * foreign value-added in demand: 0.66 (0.915) ; 0.65 (0.799)
  - Euro Area output gap:
    - Pegged * foreign value-added in demand: 0.82 (0.137)***
    - Other arrangement * foreign value-added in demand: 0.44 (0.134)***
    - Free floating * foreign value-added in demand: 0.01 (0.183)
  - Euro Area unemployment gap:
    - Pegged * foreign value-added in demand: -2.17 (0.362)***
    - Other arrangement * foreign value-added in demand: -1.07 (0.349)***
    - Free floating * foreign value-added in demand: -0.17 (0.482)
  - Euro Area instrumented headline inflation:
    - Pegged * foreign value-added in demand: 1.24 (0.239)***
    - Other arrangement * foreign value-added in demand: 0.54 (0.233)**
    - Free floating * foreign value-added in demand: 0.03 (0.302)
- Appendix Table 3: System 2SLS and 3SLS Estimation of Preferred Regression Specification — reported summary numbers:
  - 38 per country
  - System 2SLS: 0.83
  - System 3SLS: 0.82

### Definitions of explanatory variables (Appendix II)
- Headline inflation — calculated using the Harmonized Indices of Consumer Prices (HICP) published by Eurostat.
- Expected inflation — proxied by mean forecasts of average annual inflation two-years ahead published by Consensus Economics.
- World commodity price inflation — IMF’s World Economic Outlook (WEO) world oil and food price indices in US dollars; interacted with weights of energy and food in consumer baskets.
- Exchange rate appreciation/depreciation — calculated using the nominal effective exchange rates published by the IMF.
- Unemployment rate gap — cyclical unemployment rate extracted with the Baxter-King bandpass filter from data from Haver and national sources. Notes on Baxter-King filter: based on an 11-quarter centered moving average and business-cycle periodicity between 6 quarters and eight years (32 quarters). Dataset augmented with Fund staff forecasts through end-2016.
- Contribution of taxes and administered prices to headline inflation — combined contribution calculated with HICP data published by Eurostat.
- Euro Area price pressures — preferred measure: HICP core inflation (stripped of direct, first-round effects of commodity price changes). Alternative proxies used: euro area output gap, cyclical unemployment rate, instrumented headline inflation.
- Exchange rate regime dummies — based on IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions (IMF, 2013b).
- All variables are expressed in 12-month growth rates in percent, except interaction terms which are ratios.
- Share of foreign value added in domestic demand — calculated using OECD-WTO’s Trade in Value Added dataset as an average of 2005 and 2009 values. For Croatia, not available; approximated by the average of Poland and Romania due to similar import-to-GDP ratios.

*Source: _wp14191 - Appendix Figure 2. Nominal Effective Exchange Rate*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14191.pdf_