## 1. Real Appreciation in the MQ vs. the Rest of the Euro Area, 1998–2006

## Source details

**Canonical URL:** [1. Real Appreciation in the MQ vs. the Rest of the Euro Area, 1998–2006](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08240.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08240.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08240.pdf.json)

---

### Methodology: Heterogeneous-Product Approach (HPA) versus Representative-Product Approach (RPA)
- HPA identifies international competitors at the market level defined by product type (4-digit ISIC for goods, 2-digit ISIC for services), rather than by aggregate partner-country trade.
- RPA treats a single aggregated sector (a representative product category) in each destination country as the market and implicitly assumes all exporters to that destination compete with each other.
- HPA allows aggregation of market-level REER indices into a country-level REER, thereby capturing product heterogeneity and more precisely identifying direct international competitors and their weights.
- HPA incorporates competition from local producers at the destination market for each product type.
- The paper considered and rejected an approach based on cross elasticities of substitution across 4-digit industries due to measurement error in volume indices and potential identification problems from varying intensities of monopolistic competition.

### REER construction and key equations (HPA)
- Index of relative cost competitiveness between country i and the ROW:
  - lnRi,t = lnRi,T + sum from j=T+1 to t of ∆Ωi,j.
  - ∆ denotes the natural logarithmic difference between t and t−1.
- Country-level change in relative cost competitiveness:
  - ∆Ωi,t = sum over all (g,d) of βi,g,d,t−1 · ∆θi,g,d,t.
  - βi,g,d,t−1 = Si,g,d,t−1 / sum over all (g,d) Si,g,d,t−1, where Sc,g,d,t represents sales by country c of product g in destination d.
- Market-level change in relative cost competitiveness:
  - ∆θi,g,d,t = ∆Pi,g,t − sum over all c≠i αi,c,g,d,t−1 · (∆Pc,g,t + ∆Ec,i,t).
  - Pi,g,t and Pc,g,t are industry-specific cost variables in local currency; Ec,i,t is units of country i’s currency per unit of country c’s currency.
- Competitor weights and adjustment to satisfy the "10 percent property":
  - γc,g,d,t−1 = Sc,g,d,t−1 / sum over all c Sc,g,d,t−1.
  - Adjusted weighting: βi,g,d,t−1 · αi,c,g,d,t−1 = [βi,g,d,t−1 · γc,g,d,t−1] / [sum over all (c≠i,g,d) βi,g,d,t−1 · γc,g,d,t−1].
  - The adjustment ensures competitors’ weights sum to one and that if all competitors depreciate by 10 percent, country i’s REER appreciates by 10 percent.

### Data coverage: goods and services
- Goods markets: 4-digit ISIC Rev.3; 144 activity classes across 200 countries; period 1998-2005; coverage of bilateral trade in goods: above 90 percent for all MQ countries.
- Services markets: 2-digit ISIC / EBOPS 9 categories; 100 countries; period 1999-2004; structure of competitors in 1998 and 2005 extrapolated from 1999 and 2004.
- Services exports share of total exports:
  - Greece: 65 percent
  - Italy: 19 percent
  - Portugal: 27 percent
  - Spain: 31 percent
- Coverage of bilateral trade in services (average, MQ):
  - Greece: 89 percent of total services exports
  - Spain: 59 percent of total services exports
  - (Other MQ country coverage values reported in text.)
- Local production (goods) data:
  - Source: United Nations IDSB (4-digit ISIC Rev.3): 127 manufacturing commodities, 78 countries; extended using Eurostat Annual Enterprise Statistics and other methods.
  - Coverage for local production of goods (share of exports represented by destination markets with constructible local consumption data):
    - Greece: 50 percent
    - Portugal: 70 percent
    - Italy: 60 percent
    - Spain: 66 percent
- Local production (services) data:
  - Source: EU KLEMS Growth and Productivity Accounts; extended using OECD-STAN and GDP growth rates.
  - Coverage for local production of services:
    - Greece: 85 percent
    - Italy: 70 percent
    - Portugal: 83 percent
    - Spain: 94 percent
- Choice of cost competitiveness measure:
  - Main cost measure: unit labor cost (ULC).
  - Manufacturing ULC (country-level) from OECD Analytic Database and WEO; available for 38 countries.
  - Industry-level ULC at 2-digit level from EU KLEMS; covers 28 countries.

### Main quantitative findings: impact on measured REER gap, 1998–2006
- Allowing for product heterogeneity (HPA) and including services exports reduces the observed REER gap between the Mediterranean Quartet (MQ: Greece, Italy, Portugal, Spain) and the rest of the euro area relative to standard RPA-based estimates.
- Magnitude of reduction in measured real appreciation from 1998 to 2006 attributable to HPA and services (combined effect, GS vs 1G, DD estimator):
  - Greece: -2 percent
  - Italy: -2.8 percent
  - Portugal: -2.4 percent
  - Spain: -2.3 percent
- HPA vs RPA (goods only, 2006, 1998 base):
  - Portugal, Italy, Spain: REERs less appreciated in the range of 2 percent under HPA relative to RPA.
  - Greece: REER difference on the order of 7 percent (RPA more appreciated than HPA since 1998).
- Goods versus services (DD differences):
  - Italy: -3.4 percent (services appreciated less than goods)
  - Spain: -0.9 percent
  - Greece: services appreciated more than goods; DD difference on the order of 7 percent.

### Robustness checks and sensitivity
- RPA implementation consistent with Bayoumi et al. (2006 and 2005); comparison with IMF/WEO RPA-based estimates (R_IMF) shows DD estimator results within the ±1 percent range.
- Sample sensitivity:
  - Base sample includes 38 countries; comparison with 27-country sample (R_G27) yields DD estimator results within the ±1 percent range.
- Domestic market competition:
  - Including internal market information (R_GDM) yields marginal effects in the range of ±1 percent; Greece excluded due to insufficient data.
- Outlier sensitivity:
  - Eliminating the 0.5 percent tails of the distribution of annual ULC growth rates since 1998 did not substantially change results.
- Sectoral (differentiated) ULC:
  - Allowing heterogeneous sectoral cost dynamics (R_Gd vs R_G) produces absolute differences between 2 percent and 6 percent.
  - Results exploratory due to data limitations; recalculation using limited R_Gd dataset keeps contrasts between R_G and R_1G within ±1 percent.

### Structure and profile of international competitors
- Overall exposure (2005 snapshots):
  - For goods: advanced and emerging economies represent on average 95 percent of competition (except Greece 92 percent).
  - For services: advanced and emerging economies represent on average 98 percent of competition.
- Emerging economies’ share in goods (2005):
  - Spain: 14 percent
  - Italy: 19 percent
  - Portugal: 19 percent
  - Greece: 22 percent
  - China accounts for at least half of the increase in the importance of emerging economies.
- Euro area share of competition in goods (2005):
  - Spain and Portugal: 59 percent
  - Italy: 49 percent
  - Greece: 47 percent
  - Trend since 1998: declining by 1 percentage point for all MQ countries.
- Change in the share of competition from emerging economies, 1998 to 2005:
  - Greece: increase 3.4 percentage points
  - Italy: increase 4.5 percentage points
  - Portugal: increase 6.3 percentage points
  - Spain: increase 3.3 percentage points
  - France: increase 4.3 percentage points
  - Germany: increase 5.9 percentage points
- Change in percentage points due to China (1998 to 2005):
  - Greece: 3.5%
  - Italy: 3.4%
  - Portugal: 3.4%
  - Spain: 1.8%
  - France: 2.3%
  - Germany: 2.6%
- China as competitor, 2005 (shares):
  - High-tech sectors: Greece 1.43%, Italy 2.16%, Portugal 2.22%, Spain 1.36%, France 2.12%, Germany 2.64%
  - Low-tech sectors: Greece 4.17%, Italy 3.76%, Portugal 3.15%, Spain 1.79%, France 1.64%, Germany 1.18%
- Euro area share of competition in services:
  - Portugal: 60 percent
  - Italy: 50 percent
  - Spain: 48 percent
  - Greece: 37 percent
  - Trend since 1998: nil for all except Spain (decline of 5 percentage points)
- Top competitors in goods (2005, selected country shares):
  - Greece top competitors: Italy 11.84%, Germany 11.50%, France 7.13%, US 6.54%, UK 6.14%, China 5.76%, Spain 4.51%, Belgium 4.15%, Netherlands 3.93%, Turkey 3.59%.
  - Italy top competitors: Germany 18.63%, France 11.60%, US 9.72%, Spain 6.26%, UK 6.25%, China 5.95%, Japan 4.33%, Belgium 3.71%, Netherlands 3.23%, Austria 2.23%.
  - Spain top competitors (selected shares): Spain 16.92%, Germany 16.46%, Italy 9.98%, US 7.23%, Belgium 4.29%, Netherlands 4.06%, Portugal 3.95%, Japan 3.37%, China 3.23%.

### Conceptual implications
- Weighting by competitive presence in destination markets (HPA) better captures interrelationships in cost competitiveness than weighting by bilateral trade partnerships alone.
- HPA captures competition among exporters who sell to a common destination even if bilateral trade between them is nil.
- Disaggregation prevents conflating exporters of non-competing products (e.g., textiles versus cars) as direct competitors when they do not compete in the same product markets.

### Additional empirical observations and implications
- The micro-based methodology enables quantitative assessment of each country’s profile of competitors, indicating exposure to emerging competitors like China and exposure to changes in competitors’ cost structures (e.g., wage moderation in Germany, nominal exchange rate depreciation in the USA).
- Findings indicate the bulk of competition for the MQ still comes from advanced economies, especially the euro area; Spain and Portugal are more exposed to euro area competition and therefore less exposed to changes in the value of the euro.
- Since the late 1990s, emerging economies—particularly China—have grown in importance in the goods sector across both high- and low-technology sectors.

### Appendix Table A1: Availability of Unit Labor Cost Data (notes and classification)
- Industry-level ULC computed as compensation of employees per hour worked divided by real gross value added per hour worked; compensation per hour from compensation of employees divided by total hours worked by employees.
- Global goods defined per Rauch (1999) classifications mapped from SITC Rev.2 to ISIC Rev.3 using UN correspondence tables; goods with average value in [1,2) at ISIC 4-digit defined as globally traded.
- Both conservative and liberal Rauch classifications constructed; liberal classification (maximizing number of global goods) used.
- Country-level ULC compiled from OECD and IMF WEO; differentiated 2-digit sector ULC from EU KLEMS.
- Example ISIC entries and liberal/conservative/global markers are listed for numerous activities in the Appendix table.

*Source: Excerpt from IMF working paper content (_wp08240).*

### 1. Real Appreciation in the MQ vs. the Rest of the Euro Area, 1998–2006................................21

### 1. Real Appreciation in the MQ vs. the Rest of the Euro Area, 1998–2006................................21

### Methodology: Heterogeneous-Product Approach (HPA) versus Representative-Product Approach (RPA)
- HPA identifies international competitors at the market level defined by product type (4-digit ISIC for goods, 2-digit ISIC for services), rather than by aggregate partner-country trade.
- RPA treats a single aggregated sector (a representative product category) in each destination country as the market and implicitly assumes all exporters to that destination compete with each other.
- HPA allows aggregation of market-level REER indices into a country-level REER, thereby capturing product heterogeneity and more precisely identifying direct international competitors and their weights.
- HPA also incorporates competition from local producers at the destination market for each product type.
- The paper considered and rejected an approach based on cross elasticities of substitution across 4-digit industries due to measurement error in volume indices and potential identification problems from varying intensities of monopolistic competition.

### Data coverage: goods and services
- Markets are defined at 4-digit ISIC categories for goods and 2-digit ISIC categories for services.
- Services exports share of total exports:
  - Greece: 65 percent
  - Italy: 19 percent
  - Portugal: 27 percent
  - Spain: 31 percent
- Coverage of bilateral trade in services (average): ranges from 89 percent of total services exports for Greece to 59 percent for Spain.
- Coverage for goods: above 90 percent for all MQ countries.

### Main findings: impact on measured REER gap, 1998–2006
- Allowing for product heterogeneity (HPA) and including services exports reduces the observed REER gap between the Mediterranean Quartet (MQ: Greece, Italy, Portugal, Spain) and the rest of the euro area relative to standard RPA-based estimates.
- Magnitude of reduction in measured real appreciation from 1998 to 2006 attributable to HPA and services:
  - Greece: 2.0 percent
  - Italy: 2.8 percent
  - Portugal: 2.4 percent
  - Spain: 2.? percent
  - Note: exact value for Spain beyond "2.4 percent for Portugal, and" is not fully present in the supplied excerpt.

### Conceptual implications
- Weighting by competitive presence in destination markets (as in HPA) better captures interrelationships in cost competitiveness than weighting by bilateral trade partnerships alone.
- Example logic: two exporters (A and B) that sell primarily to a common third country (C) are competitors even if bilateral trade between A and B is nil; RPA may fail to capture that, while HPA will capture competition at the product-market level.
- Disaggregation prevents conflating non-competing product exporters (e.g., textiles versus cars) as direct competitors when they do not vie in the same product markets in the destination country.

*Source: IMF working paper excerpt, "1. Real Appreciation in the MQ vs. the Rest of the Euro Area, 1998–2006" (excerpted content).*

### 2.3 percent for Spain. These figures are based on difference-in-difference estimates that control for the results

### _wp08240 - 2.3 percent for Spain. These figures are based on difference-in-difference estimates that control for the results

### Key findings and robustness
- Estimated effect for Spain: 2.3 percent. These figures are based on difference-in-difference estimates that control for the results obtained for the rest of the euro area countries using the same methodology.
- Results obtained under the RPA are consistent with the ones reported using the methodology presented in Bayoumi et al. (2006 and 2005), the closest methodology to the one presented in this paper that uses RPA.
- Main cost measure used: unit labor cost (ULC) at the country-level. The paper explores differentiated ULC measures at the 2-digit level as an exploratory exercise.
- Caveats on differentiated ULC:
  - Sample with differentiated ULC series: 28 countries (more restricted than aggregated ULC sample).
  - Sample with aggregated ULC series: 38 countries.
  - Time span for differentiated ULC is one year shorter than aggregate ULC data.
  - Industry-level ULC series may be more volatile and may magnify known problems of the ULC indicator as a measure of cost competitiveness.

### Methodology overview (Heterogeneous Products Approach — HPA)
- REER index construction:
  - Index of relative cost competitiveness between country i and the rest of the world (ROW) denoted Ri,t, constructed using a geometrical Laspeyres index and chain link methodology.
  - lnRi,t = lnRi,T + sum from j=T+1 to t of ∆Ωi,j (equation 1).
  - ∆ denotes the natural logarithmic difference between t and t−1 throughout the paper.
- Country-level change in relative cost competitiveness:
  - ∆Ωi,t = sum over all (g,d) of βi,g,d,t−1 · ∆θi,g,d,t (equation 2).
  - βi,g,d,t−1 = Si,g,d,t−1 / sum over all (g,d) Si,g,d,t−1 (equation 3), where Sc,g,d,t represents sales by country c of product g in destination d.
- Market-level change in relative cost competitiveness:
  - ∆θi,g,d,t = ∆Pi,g,t − sum over all c≠i αi,c,g,d,t−1 · (∆Pc,g,t + ∆Ec,i,t) (equation 4).
  - Pi,g,t and Pc,g,t are the cost variables (industry-specific, local currency). Ec,i,t is the exchange rate defined as units of country i’s currency per unit of country c’s currency.
- Competitor weights and adjustments:
  - Market participation (share) of country c in market g,d: γc,g,d,t−1 = Sc,g,d,t−1 / sum over all c Sc,g,d,t−1 (equation 5).
  - Directly using γ would make the sum of competitors’ weights less than one (since c = i is excluded), violating a desired property.
  - Required property (the "10 percent property"): ceteris paribus, if all competitors of country i depreciate their currency by 10 percent, then country i’s REER appreciates by 10 percent.
  - Proposed adjusted weighting:
    - βi,g,d,t−1 · αi,c,g,d,t−1 = [βi,g,d,t−1 · γc,g,d,t−1] / [sum over all (c≠i,g,d) βi,g,d,t−1 · γc,g,d,t−1] (equation 6).
    - This adjusted measure satisfies that the weights of all competitors sum to one, satisfies the 10 percent property, and avoids overstating or understating relative importance of competitors.

### Aggregation choices, global goods, local consumption, and RPA
- Two elements incorporated by the literature and by the generic approach:
  - Global goods: goods whose markets are better defined at the world level (e.g., commodities such as copper). These are modeled by defining an artificial destination d = world for goods classified as global goods.
  - Local consumption of local production: competition that local producers represent at exports’ destination; proxied by the difference between local production of a good and the exports of that good from that destination to the rest of the world. Modeled by allowing d = c (competitor c competing in market g, d = c).
- RPA (Representative Partner Approach):
  - REER under the RPA is estimated to study the marginal effect of the HPA by redefining all goods into a single good g = ̄g.
  - Following Bayoumi et al. (2006 and 2005), global goods are treated separately under the RPA.
- Aggregated cost measures:
  - Where differentiated sectoral cost measures cannot be modeled due to data limitations, the generic approach can use aggregated measures by setting ∆Pc,g,t = ∆Pc,ĝ,t for all g such that g ⊂ ĝ, where ĝ refers to an aggregated sector.

### Data sources, coverage, and construction
- Goods trade data:
  - Source: United Nations COMTRADE.
  - Includes 144 different activity classes of goods (4-digit ISIC Rev.3) across 200 countries over the period 1998-2005.
  - Coverage of bilateral trade data for goods: all above 90 percent for the MQ.
- Services trade data:
  - Source: OECD Statistics of International Trade in Services.
  - Includes 9 categories of services according to EBOPS across 100 countries over the period 1999-2004.
  - Structure of competitors in 1998 and 2005 extrapolated from 1999 and 2004, respectively.
  - Average coverage of bilateral trade data for the MQ for services: ranges from 86 percent of total exports of services for Greece to 59 percent for Spain.
- Local production (goods) data:
  - Main source: United Nations Industrial Demand-Supply Balance database (IDSB) at 4-digit ISIC Rev.3 level: 127 manufacturing commodities and 78 countries.
  - Extended using:
    - (i) annual growth rates of output from Eurostat’s Annual Enterprise Statistics (4-digit NACE Rev.1.1 production data),
    - (ii) observed ratio between sectoral output and aggregate manufacturing output in Eurostat’s Annual Enterprise Statistics,
    - (iii) observed growth rate of output production in total manufacturing to extend series for a maximum window of three years (or value added growth in total manufacturing if not available).
  - Only series with complete data for the period 1998 to 2005 (original or estimated) are considered.
  - Coverage for local production of goods (share of exports represented by destination markets for which local consumption data can be constructed):
    - Greece: 50 percent
    - Portugal: 70 percent
    - Italy: 60 percent
    - Spain: 66 percent
- Local production (services) data:
  - Source: EU KLEMS Growth and Productivity Accounts database (reports data for EU25 countries, Australia, Japan, and the USA until 2005).
  - Production data for royalties and license fees not considered due to correspondence issues between EBOPS and ISIC Rev.3.
  - Series extended using OECD-STAN production data and GDP growth rates.
  - Only series with complete time series data (original or estimated) are considered.
  - Coverage for local production of services:
    - Greece: 85 percent
    - Italy: 70 percent
    - Portugal: 83 percent
    - Spain: 94 percent
- Choice of cost competitiveness measure:
  - Focus on REER measures proxying cost competitiveness using the ULC rather than CPI or PPI.
  - Advantages of ULC: incorporates changes in productivity and allows inclusion of dynamics such as the Balassa-Samuelson effect, innovation, and structural reforms.
  - Manufacturing ULC used as aggregate ULC at country-level, obtained from the OECD Analytic Database and WEO database; available for 38 countries.
  - Industry-level ULC at 2-digit level computed using the EU KLEMS database; covers 28 countries.

### Additional empirical observations reported in the text
- The micro-based methodology enables quantitative assessment of each country’s profile of competitors, indicating exposure to emerging competitors like China and exposure to changes in competitors’ cost structures (e.g., wage moderation in Germany, nominal exchange rate depreciation in the USA).
- Findings indicate that the bulk of competition for the MQ still comes from advanced economies, especially the euro area:
  - Spain and Portugal are more exposed to euro area competition and therefore less exposed to changes in the value of the euro.
- Since the late 1990s, emerging economies—particularly China—have grown in importance in the goods sector across both high- and low-technology sectors.

*Source: Excerpt from the provided IMF working paper content._wp08240 - 2.3 percent for Spain. These figures are based on difference-in-difference estimates that control for the results*

### 2005.   Table  A1  in  Appendix  details  the  different  samples.   The  industry-level  ULC  is  computed  as  the

### _wp08240 - 2005. Table A1 in Appendix details the different samples. The industry-level ULC is computed as the

### Methodology and definitions
- Industry-level ULC is computed as the ratio of the compensation of employees per hour worked to real gross value added per hour worked. The compensation of employees per hour worked is obtained from the ratio of compensation of employees to total hours worked by employees.
- Global goods: goods whose prices are quoted on organized world exchanges as defined by Rauch (1999). Rauch (1999) classifies goods at the 4-digit SITC Rev.2 level into commodities (priced on organized exchanges), reference-priced goods (listed in trade publications but not traded on organized exchanges), and differentiated goods (no reference price).
- Mapping to ISIC Rev.3: UN correspondence tables used to associate SITC Rev.2 codes with ISIC Rev.3 codes. Each ISIC Rev.3 code assigned an average value; a good is defined as globally traded at the 4-digit ISIC Rev.3 level if the average lies in the interval [1,2). Goods with values in [2,3] are not considered globally traded.
- Rauch (1999) “conservative” and “liberal” classifications are both constructed; the liberal classification (maximizing number of global goods) is used in the analysis following Jensen (2006).
- Annual average nominal exchange rates are obtained from the IFS database.
- REER estimators and contrasts:
  - R_G: REER for goods under the Heterogeneous-Product Approach (HPA).
  - R_1G: REER for goods under the Representative-Product Approach (RPA).
  - R_S: REER for services (HPA).
  - R_GS: REER for goods and services combined.
  - R_Gd: REER for goods under HPA with differentiated cost measures (sectoral ULC).
- Difference in levels (Level) measured as: Level = ∆_t% R_Gi − ∆_t% R_1Gi (equation 7).
- Difference-in-difference (DD) estimator measured as: DD = [∆_t% R_Gi − ∆_t% R_GEA] − [∆_t% R_1Gi − ∆_t% R_1GEA] (equation 8). The DD estimator is the preferred estimator.

### Main quantitative findings: sensitivity of REER to HPA and inclusion of services
- HPA vs RPA (goods):
  - Portugal’s, Italy’s, and Spain’s REERs are less appreciated in the range of 2 percent in 2006 (1998 base) under the HPA relative to the RPA.
  - Greece’s REER difference is on the order of 7 percent (RPA more appreciated than HPA since 1998).
  - Example interpretation: under the DD estimator, the REER under the RPA in Greece is 7 percent more appreciated than the HPA-based model suggests (since 1998).
- Goods versus services:
  - For MQ countries (except Greece), the services component of the REER has appreciated less than the goods component. DD differences range from -3.4 percent for Italy to -0.9 percent for Spain.
  - For Greece, services appreciated more than goods; DD difference on the order of 7 percent.
- Combined goods and services (R_GS) versus RPA (R_1G):
  - The combined effect of HPA and inclusion of services yields marginal effects on REER on the order of -2 percent to -3 percent for MQ countries:
    - -2 percent for Greece
    - -2.3 percent for Spain
    - -2.4 percent for Portugal
    - -2.8 percent for Italy

### Robustness checks and comparative validation
- Comparison with IMF/WEO RPA-based estimates (R_IMF) shows DD estimator results within the ±1 percent range, indicating similarity between the paper’s RPA implementation and existing methodologies.
- Sample sensitivity:
  - Base sample differs from R_IMF’s 27-country sample by inclusion of 11 additional emerging countries (one exception: China appears in both).
  - R_G estimated with 27-country sample (R_G27) versus 38-country sample: DD estimator results are within the ±1 percent range.
- Domestic market competition:
  - Including available information on internal markets (R_GDM) yields marginal effects with differences in the range of ±1 percent.
  - Greece excluded from this test due to insufficient disaggregated production data.
- Outlier sensitivity (sectoral ULC):
  - Eliminating the 0.5 percent tails of the distribution of annual ULC growth rates since 1998 did not substantially change results.

### Sensitivity to heterogeneous sectoral cost dynamics (differentiated ULC by sector)
- Comparing aggregated ULC REER (R_G) with differentiated-sector ULC REER (R_Gd) indicates higher sensitivity when allowing heterogeneous cost dynamics across sectors.
  - Absolute differences between R_G and R_Gd range between 2 percent and 6 percent (Table 12).
  - These results are exploratory and should be interpreted with caution due to data limitations.
- Recalculation using limited dataset available for R_Gd yields that contrasts between R_G and R_1G remain within the ±1 percent range.

### Difference in structure of competitors
- Definition: λ_i,t aggregates absolute differences in weights assigned to each competitor under HPA versus RPA (equation 9).
- Findings:
  - Greece presents the highest λ_i,t and the highest sensitivity to the HPA.
  - Portugal has the largest λ_i,t among remaining countries but does not present the highest REER sensitivity.
  - Comparing HPA to partners-approach yields a difference (λ_p) two to three times the size of λ_i,t, indicating partners-approach deviates substantially from the paper’s competitor concept.

### Profile of international competitors (exposure and composition)
- Overall exposure:
  - For goods: advanced and emerging economies represent on average 95 percent of competition (except Greece 92 percent).
  - For services: advanced and emerging economies represent on average 98 percent of competition.
- Emerging economies’ share in goods (2005):
  - Spain: 14 percent
  - Italy: 19 percent
  - Portugal: 19 percent
  - Greece: 22 percent
  - China accounts for at least half of the increase in the importance of emerging economies.
- Euro area share of competition in goods:
  - Spain and Portugal: 59 percent
  - Italy: 49 percent
  - Greece: 47 percent
  - Trend since 1998: declining by 1 percentage point for all countries.
- Sectoral exposure:
  - MQ countries compete more with China in low-technology sectors, though China’s importance in high-technology sectors is growing.
  - France and Germany compete more strongly with China in high-technology sectors than do MQ countries.
- Services composition (1999 to 2004 shifts toward emerging economies):
  - Greece: shift of 3 percent
  - Italy: shift of 3 percent
  - Portugal: shift of 2 percent
  - Spain: shift of 1 percent
- Euro area share of competition in services:
  - Portugal: 60 percent
  - Italy: 50 percent
  - Spain: 48 percent
  - Greece: 37 percent
  - Trend since 1998: nil for all countries except Spain (decline of 5 percentage points).
- Top competitors lists are provided in the source (Tables 16 and 17 for goods and services respectively).

### Conclusion: implications for REER measurement
- Incorporating product heterogeneity (HPA) and services exports refines REER measurement and identification of direct competitors.
- The combined effect of micro-based competitor structure and inclusion of services implies a modest lower real appreciation from 1998 to 2006 on the order of 2-3 percent for MQ countries:
  - 2.0 percent for Greece (reported in conclusion)
  - 2.8 percent for Italy (reported in conclusion)
  - 2.4 percent for Portugal (reported in conclusion)
- Sectoral cost heterogeneity can increase sensitivity of REER estimates, with potential absolute differences up to 6 percent in exploratory sectoral analyses.

*Source: _wp08240 - 2005. Table A1 in Appendix details the different samples. The industry-level ULC is computed as the*

### 2.3 percent for Spain. These estimates are based on a difference-in-difference estimator that controls for the

### _wp08240 - 2.3 percent for Spain. These estimates are based on a difference-in-difference estimator that controls for the

### Methodology and Definitions
- ULC-based REER index defined à la IMF (higher means more appreciated); the base year is 1998.
- EA-8 refers to Austria, Belgium, France, Finland, Germany, Ireland, Luxembourg, and Netherlands.
- The REER for the EA-8 is estimated aggregating each countries' REER and weighting by their total exports of goods.
- All countries adopted the euro on January 1, 1999, except for Greece, which adopted the euro on January 1, 2001.
- The paper uses a heterogeneous-product approach to define competitors (disaggregated goods and services categories), contrasted with the representative-product approach.
- Difference-in-difference estimator is used to measure net appreciation differentials controlling for effects observed in the rest of 11 euro area countries.

### Key Findings: Real Appreciation and Competitiveness
- The bulk of competition for the MQ still comes from the advanced economies, especially from the euro area.
- Emerging economies gained importance as competitors since the late 1990s, particularly China in both high- and low-technology sectors.
- Change in the share of competition from emerging economies, 1998 to 2005:
  - Greece: increase 3.4 percentage points
  - Italy: increase 4.5 percentage points
  - Portugal: increase 6.3 percentage points
  - Spain: increase 3.3 percentage points
  - France: increase 4.3 percentage points
  - Germany: increase 5.9 percentage points
- Change in percentage points due to China (1998 to 2005):
  - Greece: 3.5%
  - Italy: 3.4%
  - Portugal: 3.4%
  - Spain: 1.8%
  - France: 2.3%
  - Germany: 2.6%

### China and Germany as Competitors (selected statistics)
- China as competitor, 2005 (share in relevant sectors):
  - High-tech sectors: Greece 1.43%, Italy 2.16%, Portugal 2.22%, Spain 1.36%, France 2.12%, Germany 2.64%
  - Low-tech sectors: Greece 4.17%, Italy 3.76%, Portugal 3.15%, Spain 1.79%, France 1.64%, Germany 1.18%
- Germany as competitor, 2005 (share in relevant sectors):
  - High-tech sectors: Greece 5.01%, Italy 11.94%, Portugal 7.84%, Spain 12.82%, France 13.74%, Germany ...
  - Low-tech sectors: Greece 6.29%, Italy 6.63%, Portugal 4.61%, Spain 3.90%, France 4.95%, Germany ...

### Structure of Competitors: Aggregate shares (selected snapshots)
- Goods, Euro area share of competitors:
  - Euro area, 1998: Greece 37.4%, Italy 49.6%, Portugal 60.3%, Spain 54.4%, France 42.8%, Germany 46.0%
  - Euro area, 2004: Greece 37.2%, Italy 50.2%, Portugal 60.3%, Spain 48.1%, France 42.5%, Germany 40.7%
- Goods, Advanced economies share:
  - Advanced economies, 1998: Greece 95.9%, Italy 94.1%, Portugal 97.1%, Spain 95.8%, France 95.1%, Germany 93.6%
  - Advanced economies, 2004: Greece 92.0%, Italy 89.7%, Portugal 94.7%, Spain 93.5%, France 91.1%, Germany 87.8%
- Services, Euro area and advanced/emerging shares (selected rows):
  - Euro area, 1998: Greece 47.5%, Italy 49.7%, Portugal 60.0%, Spain 59.9%, France 48.9%, Germany 42.6%
  - Euro area, 2005: Greece 47.0%, Italy 48.6%, Portugal 58.6%, Spain 58.5%, France 48.6%, Germany 41.1%
  - Advanced economies, 1998: Greece 72.4%, Italy 81.4%, Portugal 83.7%, Spain 85.9%, France 85.7%, Germany 84.5%
  - Advanced economies, 2005: Greece 69.9%, Italy 75.7%, Portugal 77.0%, Spain 81.5%, France 80.1%, Germany 77.4%
  - Emerging economies, 1998: Greece 18.6%, Italy 14.7%, Portugal 12.7%, Spain 11.1%, France 11.7%, Germany 11.8%
  - Emerging economies, 2005: Greece 22.0%, Italy 19.2%, Portugal 19.0%, Spain 14.4%, France 15.9%, Germany 17.7%

### Main Competitors in 2005: Goods (top ranks and shares)
- Greece, top competitors (share):
  1. Italy (11.84%)
  2. Germany (11.50%)
  3. France (7.13%)
  4. US (6.54%)
  5. UK (6.14%)
  6. China (5.76%)
  7. Spain (4.51%)
  8. Belgium (4.15%)
  9. Netherlands (3.93%)
  10. Turkey (3.59%)
- Italy, top competitors (share):
  1. Germany (18.63%)
  2. France (11.60%)
  3. US (9.72%)
  4. Spain (6.26%)
  5. UK (6.25%)
  6. China (5.95%)
  7. Japan (4.33%)
  8. Belgium (3.71%)
  9. Netherlands (3.23%)
  10. Austria (2.23%)
- Spain, top competitors (share):
  1. Spain (16.92%) [table shows Spain as top for Spain]
  2. Germany (16.46%)
  3. Italy (9.98%)
  4. US (7.23%)
  5. Spain (6.49%) [duplicate entries reflect goods vs. services tables]
  6. Belgium (4.29%)
  7. Netherlands (4.06%)
  8. Portugal (3.95%)
  9. Japan (3.37%)
  10. China (3.23%)
- France and Germany entries similarly report top competitor shares (see tables for exact ranked percentages).

### Heterogeneous vs Representative-Product and Robustness Checks (selected outcomes)
- The heterogeneous-product approach (G) yields different ULC-based REER levels and net appreciation differentials compared with the representative-product approach (1G) and the IMF benchmark.
- Tables report annual "Level" (difference between growth rates) and "DD" (difference-in-difference estimate) in percentages for comparisons: G vs 1G, 1G vs IMF, G vs G27 (reduced sample), G vs GIM (including domestic market competition), S vs G (services vs goods), GS vs 1G (goods and services vs representative goods), and Gd vs G (differentiated ULC vs country-level ULC).
- Example robustness observations (selected years, percentages preserved):
  - Net appreciation differential (1G vs IMF), Spain, 2006 Level: -0.31, DD: 0.03 (results presented in percentages).
  - Joint marginal effect of heterogeneous-product including services (GS vs 1G), Greece 2005 Level: 5.57, DD: 4.24 (results presented in percentages).
  - Marginal effect of domestic market competition (G vs GIM) shows sizable Level and DD values across countries and years (tables present values in percentages, e.g., Spain 2006 Level: -1.24, DD: -0.86).

### Additional Data Notes
- Technology intensity classification follows OECD (2007): high-tech comprises industries with high and medium-high intensity; low-tech comprises industries with low and medium-low intensity.
- The sample composition for differentiated ULC measures lists countries and which samples include differentiated ULC; sample sizes referenced include 38 countries, 27 countries, and 28 countries with explicit "Yes"/"No" entries by country.

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08240.pdf*

### Appendix Table A1. Availability of Unit Labor Cost Data

### Appendix Table A1. Availability of Unit Labor Cost Data

### Notes
- Country-level ULC data compiled from the OECD database and IMF World Economic Outlook database.
- Differentiated 2-digit sector level ULC data compiled from EU KLEMS.
- We refer to globally traded goods as goods whose prices are quoted on organized world exchanges as defined by Rauch (1999). These goods are characterized as commodities for which a more appropriate definition of market is at the world level.
- Rauch distinguishes between conservative and liberal classifications. We use the liberal classification that maximizes the number of global goods.
- For comparison reasons, we identify an additional list of global goods (denoted by IMF) in line with Bayoumi et al. (2005).

### Country-level and Sector-level Availability (ISIC Rev.3 activity list and availability markers)
- ISIC Rev.3 — Activity Description — IMF — (4-digit)ConservativeLiberal
- 0111 Growing of cereals and other crops n.e.c. — Global Global Global
- 0112 Growing of vegetables, horticultural specialties and nursery products — // Global
- 0113 Growing of fruit, nuts, beverage and spice crops — Global Global Global
- 0121 Farming (cattle, sheep, goats, horses, asses, mules and hinnies; dairy) — Global Global Global
- 0122 Other animal farming; production of animal products n.e.c. — // Global
- 0200 Forestry, logging and related service activities — // Global
- 0500 Fishing operations — // Global
- 1110 Extraction of crude petroleum and natural gas — / Global /
- 1200 Mining of uranium and thorium ores — / Global Global
- 1310 Mining of iron ores — Global Global Global
- 1320 Mining of non-ferrous metal ores, except uranium and thorium ores — / Global Global
- 1410 Quarrying of stone, sand and clay — // Global
- 1421 Mining of chemical and fertilizer minerals — // Global
- 1422 Extraction of salt — // Global
- 1429 Other mining and quarrying n.e.c. — // Global
- 1511 Production, processing and preserving of meat and meat products — Global Global Global
- 1512 Processing and preserving of fish and fish products — // Global
- 1513 Processing and preserving of fruit and vegetables — // Global
- 1514 Manufacture of vegetable and animal oils and fats — Global Global Global
- 1520 Manufacture of dairy products — / Global Global
- 1531 Manufacture of grain mill products — // Global
- 1532 Manufacture of starches and starch products — // Global
- 1533 Manufacture of prepared animal feeds — // Global
- 1541 Manufacture of bakery products — // Global
- 1542 Manufacture of sugar — Global Global Global
- 1543 Manufacture of cocoa, chocolate and sugar confectionery — // Global
- 1544 Manufacture of farinaceous products (macaroni and similar) — // Global
- 1549 Manufacture of other food products n.e.c. — // Global
- 1551 Distilling, rectifying and blending of spirits — // Global
- 1552 Manufacture of wines — // Global
- 1553 Manufacture of malt liquors and malt — // Global
- 1554 Manufacture of soft drinks; production of mineral waters — // Global
- 1600 Manufacture of tobacco products — // Global
- 2010 Sawmilling and planing of wood — // Global
- 2411 Manufacture of basic chemicals, exc. fertilizers & nitrogen compounds — Global Global /
- 2412 Manufacture of basic precious and non-ferrous metals — / Global /
- 2720 Manufacture of basic precious and non-ferrous metals — Global Global Global
- 9302 Hairdressing and other beauty treatment — // Global

### Key statistics and totals
- Total91435

### Classification and methodology
- Rauch Classification is used to identify globally traded goods; both conservative and liberal classifications exist.
- The liberal classification is used in this table to maximize the number of global goods.
- An additional IMF list of global goods is identified for comparison, in line with Bayoumi et al. (2005).

*Source: _wp08240 - Appendix Table A1. Availability of Unit Labor Cost Data*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08240.pdf_