## 1. Per Capita GDP Compared with Other European Small Economies and Microstates

## Source details

**Canonical URL:** [1. Per Capita GDP Compared with Other European Small Economies and Microstates](https://www.imf.org/-/media/files/publications/cr/2025/english/1smrea2025002-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2025/english/1smrea2025002-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2025/english/1smrea2025002-source-pdf.pdf.json)

---

### Introduction and analytical approach
- Purpose:
  - Assess economic and export diversification in San Marino using sectoral Gross Value Added (GVA) data and export diversification metrics—Herfindahl-Hirschman Index (HHI), Entropy Index (EI), and Export Concentration Ratio (CR)—derived from mirror trade data.
- Core results (preview):
  - Aggregate sectoral data show concentration in manufacturing and wholesale trade.
  - Disaggregated product-level export data indicate San Marino’s export structure is well diversified given country size and development level, despite recent slight decreases in diversification.
  - Analysis confirms a U-shaped relationship between diversification and income; high-income economies like San Marino may not gain further from additional diversification.
- Policy implication emphasized:
  - Prioritize horizontal structural policies to support high-value-added products within competitive sectors and promote high-productivity growth sectors with economy-wide synergies (e.g., ICT), rather than sector-specific industrial policy.

### Key contextual literature points
- Economic diversification reduces vulnerability to sector-specific shocks (IMF 2014; UNCTAD, 2017) and fosters innovation and productivity through structural transformation (Herreford, Rogerson and Valentinyi, 2014; Rodrik, 2006; McMillan and Rodrik, 2011).
- Export diversification is associated with higher growth and productivity (Al-Marhubi, 2000; Gutierrez de Pineres and Ferrantino, 2000; Lerderman and Maloney, 2007; Papageorgiou et al., 2015; McIntyre et al., 2018; Henn et al, 2020; Feenstra and Kee, 2004).
- For high-income economies, further growth is often associated with specialization rather than diversification (Rodrik, 2006; Klinger and Lederman, 2004; Imbs and Wacziarg, 2003).

### Data and scope notes
- Population and scale:
  - San Marino’s population is 34 thousand.
  - San Marino’s economy accounts for 0.0016 percent of World GDP as of 2024.
- Trade data:
  - Export diversification metrics computed from mirror SITC product-level trade data (backed out from partners’ imports where direct SITC for San Marino is unavailable).

### Stylized facts: per capita and per worker GDP
- Relative income levels:
  - San Marino has a high level of per capita income compared with EU and other European microstates (except Luxembourg and Liechtenstein) during most of the 1997-2023 period.
  - Since the Global Financial Crisis (GFC), San Marino’s per capita GDP has declined relative to peers, largely due to the deleterious effect on its then-oversized financial sector.
  - GDP per worker exhibits the same broad pattern; the post-GFC decline brought per capita GDP down to the average EU level and below Canada more recently.

### Stylized facts: sectoral composition and dynamics (sector shares of total GVA, 2023 unless otherwise indicated)
- Sectoral shares:
  - Manufacturing: 34.2 percent
  - Wholesale and retail trade: 15.7 percent
  - Public administration: 13.4 percent
  - Agriculture: about 0.02 percent
  - Mining: 0.4 percent
  - Transport: about 2 percent in San Marino (compare: 4.2 percent in Luxembourg)
- Cross-country contrasts:
  - Andorra and Luxembourg have much lower manufacturing shares (only 4 percent) but much larger finance and insurance shares (13 and 25 percent, respectively) versus 4 percent in San Marino.
  - Andorra’s real estate share: 16 percent; San Marino real estate: 6 percent.
  - Luxembourg’s wholesale and retail trade: 8 percent (about half of San Marino’s).
- Sectoral changes 2015–2023:
  - Manufacturing share increased by more than 3 percentage points relative to 2015.
  - Wholesale and retail trade: +3.6 ppt.
  - ICT: +1.6 ppt.
  - Combined, manufacturing, wholesale and retail, and ICT account for about 3/4 of cumulative GVA growth since 2015 (the three sectors combined account for almost 3/4 of total GVA growth in San Marino since 2015).
  - Sectors with declining GVA shares: finance and insurance and real estate (both -1.8 ppt), professional services (-2.3 ppt), public administration (-3 ppt), arts/entertainment/other (-0.5 ppt).

### Employment, exports, and correlations with GVA
- Sectoral match and correlations:
  - PPMC between sectoral GVA and sectoral employment: about 0.95 since 2015, reaching 0.96 in 2022-2023.
  - PPMC between sectoral GVA and sectoral exports of goods: close to 0.9 during 2015-2023, reaching 0.94 in 2022-2023.
- Sectoral shares in employment and exports (2024 or latest available; exact phrasing preserved):
  - Manufacturing: about 1/3 of total employment; about 55 percent of goods exports.
  - Wholesale and retail trade: constitutes about 15 and 22 percent, respectively.
  - Public administration: 19.1 percent share of employment.
  - Together, manufacturing, wholesale and retail, and public administration employ about 2/3 of workers; manufacturing and wholesale and retail account for more than 3/4 of exports.

### Sectoral labor productivity
- Trends 2015–2022:
  - Labor productivity grew steadily in all but two sectors: agriculture and real estate.
  - The three expanding sectors (ICT, manufacturing, wholesale and retail trade) showed productivity growth.
- Productivity levels:
  - Highest in real estate, followed by finance and insurance, though both are showing declines.
  - Excluding real estate (distorted by the minuscule domestic market), the correlation between the average sectoral productivity level in 2023 and the sectoral cumulative productivity growth since 2015 is 0.6.

### Export orientation and trade openness
- Export intensity and manufacturing export propensity:
  - Ratio of total exports to gross production: 54 percent by 2022.
  - Ratio of total exports to GVA: 200 percent by 2023.
  - Manufacturing exports: exports 87 percent of its output.
- Implication:
  - Very high openness and export orientation given San Marino’s small domestic market; some sectors almost exclusively produce for export.

### Export diversification analysis: approach and metrics
- Rationale:
  - Sector-level concentration can be misleading; intra-sector product-level diversity matters.
- Indicators used:
  - Herfindahl-Hirschman Index (HHI), Entropy Index (EI), Export Concentration Ratio (CR).
- Comparator set:
  - Small European economies and microstates, augmented by Canada, Switzerland, and Mauritius.

### Main analytical findings on diversification (text summary)
- Aggregate-sector results:
  - Aggregate GVA, employment, and exports are concentrated in manufacturing and wholesale and retail sectors.
- Product-level export results:
  - Product-level export data (mirror SITC) show San Marino’s export basket is well diversified given its size and development level, though there are indications of a slight recent decrease in diversification.
- Income-diversification relationship:
  - The relationship between diversification and income is U-shaped; high-income economies like San Marino may not further benefit from more diversification.
- Trade openness nuance:
  - Trade openness lowers output volatility in sufficiently diversified economies but increases volatility in those with concentrated export baskets (Haddad et al., 2013).

### Herfindahl-Hirschman Index (HHI): definition and cross-country patterns
- Definition:
  - The HHI sums the squares of the market shares of all exported products or sectors; it weighs larger shares more heavily but considers the entire distribution of export shares. The lower the index, the less concentrated exports are.
- Cross-country patterns (small and microstate peer group, 2022):
  - Slightly positive association between the HHI and per capita GDP.
  - Clear negative relationship between the HHI and population size, suggesting small population size poses a limit to diversification.
  - Relationships hold regardless of SITC product disaggregation level (1 or 4 digits).
- San Marino’s relative position (by 2022):
  - At the regression line regarding per capita GDP and below it when considering population size, implying:
    - Export diversification is consistent with its development stage.
    - Export diversification is above what would have been predicted by its small population size.
  - Direct comparisons:
    - More diversified than Iceland and Andorra.
    - Less diversified than Malta, Luxembourg, Estonia, and Latvia.
    - At SITC 1-digit, as diversified as Mauritius.
    - At SITC 4-digit, more diversified than Cyprus and Switzerland.
  - Time trends:
    - At the 1-digit level, slight decline in diversification since 2000.
    - At the 4-digit level, decline more pronounced, especially since 2016.

### Panel regression evidence: HHI vs GDP per capita and population (2000–2022)
- Data and method:
  - Annual data from 56 OECD and European countries during 1988−2023; panel regression with country-fixed effects of HHI on (i) per capita GDP, (ii) per capita GDP squared, and (iii) population size. All variables entered in natural logs.
- Main regression findings (dependent variable Ln(HHI); sample: 2000-2022; # Countries = 56; Observations = 1788):
  - Relationship between HHI and per capita GDP is U-shaped (convex), confirmed across SITC disaggregation levels (1 to 5 digits):
    - Coefficient of Ln (GDP per capita) is always negative and highly statistically significant.
    - Coefficient of Ln (GDP per capita)^2 is always positive and highly statistically significant.
  - Implied elasticity ranges:
    - Results suggest a reduction of 1 percent in HHI for increases in per capita GDP ranging from 0.7 percent (SITC 1-digit) to 3.3 percent (SITC 5-digit).
  - Population effects:
    - For small economies and microstates there is a statistically significant negative relationship between HHI and population size.
    - In the wider country sample, the negative relationship with population disappears in most specifications; the coefficient of population is not statistically significant in all but one specification.
    - Exception: at the SITC 5-digit level a statistically significant positive HHI-population association is found.
- Selected coefficient estimates (preserved exactly):
  - Ln (GDP per capita): -0.73***, -0.84***, -1.84***, -1.64***, -2.33***, -2.04***, -1.94***, -1.93***, -3.32***, -2.45***.
  - Ln (GDP per capita^2): 0.04***, 0.04***, 0.10***, 0.08***, 0.13***, 0.11***, 0.13***, 0.13***, 0.18***, 0.13***.
  - Ln (Population) (selected reported values): -0.06, 0.12, 0.17, 0.00, 0.50***.
  - Constant (selected reported values): 2.13**, 3.58**, 6.23***, 3.47, 7.18***, 3.25, 3.33, 3.31, 10.59***, -1.11.
  - Note: * p-value < 0.1; ** p-value < 0.05; and *** p-value < 0.01.
- Interpretation for San Marino:
  - San Marino’s data cluster around the fitted regression curve for both SITC 1- and 4-digit levels, in a portion of the curve that surpassed the threshold level of GDP per capita where the relationship becomes positive.
  - Overall, San Marino is more diversified than its size would predict; for 1- and 4-digit disaggregation San Marino’s HHI is at or below the partial regression curve of HHI on population size.

### Entropy Index (EI): characteristics and San Marino’s position
- Definition and sensitivity:
  - EI captures dispersion in the distribution of export products, summing each export product’s market share multiplied by its natural log (typically presented with a minus sign so higher EI means more diversified).
  - EI treats product shares more equally and is more sensitive to the presence of very small shares than the HHI.
- Cross-country comparisons and time patterns:
  - EI-based comparisons broadly confirm HHI-based results.
  - At SITC 1- and 4-digit levels, San Marino appears:
    - More diversified than Iceland and Andorra.
    - More concentrated than Luxembourg, Estonia, and Latvia.
    - At SITC 4-digit, more diversified than Cyprus and Malta, and at par with Switzerland.
  - EI shows a decline in diversification for San Marino since the mid-2010s.
- Interpretation:
  - Similar overall pictures from HHI and EI are consistent with an evenly distributed export structure in San Marino: while some products dominate, San Marino does not display a very long tail of minor exports compared with peers.

### Export Concentration Ratio (CR-4): focused measure of top goods
- Definition and sensitivity:
  - CR-N sums the shares of the top N export products; CR-4 focuses on the top 4 export products at a given SITC level. It measures concentration at the top of the distribution and does not account for the remainder of the export basket.
- San Marino’s CR-4 findings:
  - At SITC 1-digit:
    - San Marino is currently less CR-diversified than most peers, except Andorra and, by a small margin, Iceland and Switzerland.
    - San Marino’s CR-4 has displayed a secular decline since the early 2000s at this aggregation level.
  - At SITC 4-digit:
    - The relative ranking aligns with HHI- and EI-based comparisons: San Marino is more diversified than Iceland, Andorra, Cyprus, and Switzerland by CR-4.
    - San Marino is becoming less diversified more recently at the 4-digit level as well.
- Interpretation:
  - The top 4 export products are more concentrated at an aggregated level in San Marino; reliance on a small number of key products reduces with more product disaggregation.

### Synthesis of index-specific evidence
- Aggregate-sector versus product-level:
  - At aggregate sectoral levels San Marino appears concentrated (manufacturing; wholesale and retail).
  - At more disaggregated product levels San Marino emerges as a well-diversified economy:
    - HHI indicates a relatively diversified export structure compared to peer microstates.
    - EI suggests a balanced distribution relative to peers (no pronounced long tail of minor exports).
    - CR-4 shows concentration in the top 4 goods at the aggregated 1-digit level, but this concentration disappears at the disaggregated 4-digit level.
- Policy-relevant implication:
  - While a few key products may constitute a larger share of exports in aggregated data, detailed disaggregation shows multiple products with meaningful shares. Policies should weigh aggregate-sector concentration signals against the more nuanced picture from disaggregated export-product data when designing diversification or resilience strategies.

### Resilience and recent growth developments
- Growth model shift:
  - San Marino diversified away from financial services after the Global Financial Crisis.
- Economic performance:
  - The economy has maintained positive growth following the post-pandemic successive shocks.
  - Total employment reached record highs.
- Sectoral contributors to resilience:
  - Manufacturing retained market share gained through nearshoring.
  - Tourism flourished on post-Covid regional demand boom.

### Policy implications and recommendations
- Overarching recommendation:
  - Emphasize horizontal structural policies rather than sector-specific industrial policy.
- Suggested horizontal reforms to strengthen existing productive sectors and boost economy-wide productivity:
  - Investing in both physical and digital infrastructure.
  - Increasing labor market flexibility.
  - Streamlining and modernizing the tax system.
  - Resolving banking-real estate distortions to facilitate real estate market functioning.
  - Enhancing public administration efficiency.
- Sectoral focus:
  - Support development of high-value-added products within currently competitive sectors (notably manufacturing and wholesale and retail).
  - Support high-productivity-growth sectors with economy-wide synergies—explicitly: ICT—to sustain productivity growth while maintaining resilience through diversification.
- Rationale:
  - Given San Marino’s development stage and that its per capita GDP has surpassed the diversification inflection point, horizontal policies and strengthening of existing comparative advantages are preferred over efforts to pursue more diversification per se.

### Annex I — Index definitions (formulas preserved)
- Herfindahl-Hirschman Index (HHI):
  - Formula: HHHHHH = ∑ si2 Ni=1, where si is the share of product i in total exports and N is the total number of products.
  - Interpretation: A lower HHI indicates a more diversified export base; a higher HHI indicates concentration on fewer products.
- Entropy Index:
  - Formula: EEEEEEEE = − ∑ si Ni=1 . ln si, where si is the share of product i in total exports and N is the total number of products.
  - Interpretation: A higher entropy value indicates a more diversified export structure.
- Export Concentration Ratio (CR):
  - Definition: Share of the top few products in total exports (e.g., CR-N measures the concentration of the top N export products).
  - Interpretation: A lower ratio indicates a more diversified export sector.

### Annex II — Select empirical results (preserved values)
- Panel regression: Entropy vs Country Size and Per Capita GDP, 2000-2022 (dependent variable: Entropy Index at various SITC digit levels)
  - Ln (GDP per capita): 0.47***, 0.60***, 1.41***, 1.25***, 1.99***, 1.53***, 3.11***, 2.37***, 3.27***, 3.40***
  - Ln (GDP per capita^2): -0.02***, -0.03***, -0.07***, -0.06***, -0.11***, -0.08***, -0.17***, -0.13***, -0.18***, -0.19***
  - Ln (Population): 0.07*, -0.09, -0.26***, -0.42***, 0.07
  - Constant: -0.89, -2.55**, -3.68***, -1.49, -4.94***, 1.19, -9.25***, 0.63, -10.56***, -12.31***
  - Observations 1788; # Countries 56.
- Panel regression: Export Concentration Ratio (Ln(CR)) vs Country Size and Per Capita GDP, 2000-2022
  - Ln (GDP per capita): -0.08*, -0.04, -1.23***, -0.95***, -1.70***, -1.38***, -1.46***, -1.32***, -1.31***, -0.48
  - Ln (GDP per capita^2): 0.00, 0.00, 0.07***, 0.05***, 0.10***, 0.08***, 0.09***, 0.09***, 0.07***, 0.03
  - Ln (Population): 0.02*, 0.16***, 0.19***, 0.08, 0.47***
  - Constant: 0.36, -0.21, 4.84***, 1.06, 6.23***, 1.90, 3.91***, 2.04, 3.72**, -7.27***
  - Observations 1788; # Countries 56.

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

### 1. Per Capita GDP Compared with Other European Small Economies and Microstates  ___ 4

### 1. Per Capita GDP Compared with Other European Small Economies and Microstates

### Introduction and analytical approach
- Purpose: assess economic and export diversification in San Marino, a high-income small state with a manufacturing-heavy economy, using sectoral Gross Value Added (GVA) data and export diversification metrics—Herfindahl-Hirschman Index (HHI), Entropy Index (EI), and Export Concentration Ratio (CR)—derived from mirror trade data.
- Core result preview:
  - Aggregate sectoral data show concentration in manufacturing and wholesale trade.
  - Disaggregated product-level export data indicate San Marino’s export structure is well diversified given country size and development level, despite recent slight decreases in diversification.
  - Analysis confirms a U-shaped relationship between diversification and income, implying high-income economies like San Marino may not gain further from additional diversification.
- Policy implication highlighted in the paper: prioritize horizontal structural policies to support high-value-added products within currently competitive sectors and promote high-productivity growth sectors with economy-wide synergies (e.g., ICT), rather than sector-specific industrial policy.

### Key contextual literature points (preserved terminology)
- Economic diversification reduces vulnerability to sector-specific shocks (IMF 2014; UNCTAD, 2017) and fosters innovation and productivity through structural transformation (Herreford, Rogerson and Valentinyi, 2014; Rodrik, 2006; McMillan and Rodrik, 2011).
- Export diversification is associated with higher growth and productivity (Al-Marhubi, 2000; Gutierrez de Pineres and Ferrantino, 2000; Lerderman and Maloney, 2007; Papageorgiou et al., 2015; McIntyre et al., 2018; Henn et al, 2020; Feenstra and Kee, 2004).
- For high-income economies, further growth is often associated with specialization rather than diversification (Rodrik, 2006; Klinger and Lederman, 2004; Imbs and Wacziarg, 2003).

### Data and scope notes
- San Marino’s population is 34 thousand, while its economy accounts for 0.0016 percent of World GDP as of 2024.
- Export diversification metrics are computed from mirror SITC product-level trade data (backed out from partners’ imports where direct SITC for San Marino is unavailable).

### Stylized facts: per capita and per worker GDP
- San Marino has a high level of per capita income compared with EU and other European microstates (except Luxembourg and Liechtenstein) during most of the 1997-2023 period.
- Since the Global Financial Crisis (GFC), San Marino’s per capita GDP has declined relative to peers, largely due to the deleterious effect on its then-oversized financial sector.
- GDP per worker exhibits the same broad pattern; the post-GFC decline brought per capita GDP down to the average EU level and below Canada more recently.

### Stylized facts: sectoral composition and dynamics
- Sectoral shares of total GVA (2023 unless otherwise indicated):
  - Manufacturing: 34.2 percent
  - Wholesale and retail trade: 15.7 percent
  - Public administration: 13.4 percent
  - Agriculture: about 0.02 percent
  - Mining: 0.4 percent
  - Transport: about 2 percent in San Marino (compare: 4.2 percent in Luxembourg)
- Cross-country contrasts:
  - Andorra and Luxembourg have much lower manufacturing shares (only 4 percent) but much larger finance and insurance shares (13 and 25 percent, respectively) versus 4 percent in San Marino.
  - Andorra’s real estate share: 16 percent; San Marino real estate: 6 percent.
  - Luxembourg’s wholesale and retail trade: 8 percent (about half of San Marino’s).
- Sectoral changes 2015–2023:
  - Manufacturing share increased by more than 3 percentage points relative to 2015.
  - Wholesale and retail trade: +3.6 ppt.
  - ICT: +1.6 ppt.
  - Combined, manufacturing, wholesale and retail, and ICT account for about 3/4 of cumulative GVA growth since 2015 (the three sectors combined account for almost 3/4 of total GVA growth in San Marino since 2015).
  - Sectors with declining GVA shares: finance and insurance and real estate (both -1.8 ppt), professional services (-2.3 ppt), public administration (-3 ppt), arts/entertainment/other (-0.5 ppt).

### Employment, exports, and correlations with GVA
- Sectoral match:
  - PPMC between sectoral GVA and sectoral employment: about 0.95 since 2015, reaching 0.96 in 2022-2023.
  - PPMC between sectoral GVA and sectoral exports of goods: close to 0.9 during 2015-2023, reaching 0.94 in 2022-2023.
- Sectoral shares in employment and exports (2024 or latest available):
  - Manufacturing: about 1/3 of total employment; about 55 percent of goods exports.
  - Wholesale and retail trade: constitutes about 15 and 22 percent, respectively. (Exact phrasing preserved from source.)
  - Public administration: 19.1 percent share of employment.
  - Together, manufacturing, wholesale and retail, and public administration employ about 2/3 of workers; manufacturing and wholesale and retail account for more than 3/4 of exports.

### Sectoral labor productivity
- Labor productivity is uneven across sectors and has generally been growing faster in expanding sectors.
- Between 2015 and 2022:
  - Labor productivity grew steadily in all but two sectors: agriculture and real estate.
  - The three expanding sectors (ICT, manufacturing, wholesale and retail trade) showed productivity growth.
- Productivity level:
  - Highest in real estate, followed by finance and insurance, though both are showing declines.
  - Excluding real estate (distorted by the minuscule domestic market), the correlation between the average sectoral productivity level in 2023 and the sectoral cumulative productivity growth since 2015 is 0.6.

### Export orientation and trade openness
- Export intensity and manufacturing export propensity:
  - Ratio of total exports to gross production: 54 percent by 2022.
  - Ratio of total exports to GVA: 200 percent by 2023.
  - Manufacturing exports: exports 87 percent of its output.
- Implication: very high openness and export orientation given San Marino’s small domestic market; some sectors almost exclusively produce for export.

### Export diversification analysis: approach and metrics
- Rationale: sector-level concentration can be misleading; intra-sector product-level diversity matters.
- Indicators used as proxies for economic diversification: Herfindahl-Hirschman Index (HHI), Entropy Index (EI), Export Concentration Ratio (CR).
- Cross-country comparator set: small European economies and microstates, augmented by Canada, Switzerland, and Mauritius.
- Technical Annex I discusses indicator calculations (not reproduced here).

### Main analytical findings on diversification (summary from text)
- Aggregate GVA, employment, and exports are concentrated in manufacturing and wholesale and retail sectors (sector-level concentration).
- Product-level export data (mirror SITC) show that San Marino’s export basket is well diversified given its size and development level, though there are indications of a slight recent decrease in diversification.
- The relationship between diversification and income is U-shaped; high-income economies like San Marino may not further benefit from more diversification.
- Trade openness considerations: trade openness lowers output volatility in sufficiently diversified economies but increases volatility in those with concentrated export baskets (Haddad et al., 2013).

### Policy implications and recommendations (preserved emphasis)
- Given the evidence and San Marino’s development stage, the paper recommends:
  - Emphasize horizontal structural policies rather than sector-specific industrial policy.
  - Support development of high-value-added products within currently competitive sectors (notably manufacturing and wholesale and retail).
  - Support high-productivity-growth sectors with economy-wide synergies—explicitly: ICT is highlighted as relevant—to sustain productivity growth while maintaining resilience through diversification.

*Prepared by Carlos de Resende and Jenny Lee (EUR); content excerpted from the IMF paper "Assessing Economic Diversification in San Marino" (October 31, 2025).*

### 20.      The HHI sums the squares of the market shares of all exported products or sectors. It

### Export diversification analysis: HHI, Entropy Index, and Concentration Ratios for San Marino

### Herfindahl-Hirschman Index (HHI): definition and cross-country patterns
- The HHI sums the squares of the market shares of all exported products or sectors; it weighs larger shares more heavily but considers the entire distribution of export shares. The lower the index, the less concentrated exports are.
- In the small and microstate peer group (2022):
  - There is a slightly positive association between the HHI and per capita GDP.
  - There is a clear negative relationship between the HHI and population size, suggesting that small population size poses a limit to diversification.
  - These relationships hold regardless of SITC product disaggregation level (either 1 or 4 digits).
- San Marino’s relative position:
  - By 2022, San Marino is at the regression line regarding per capita GDP and below it when considering population size, implying:
    - Export diversification is consistent with its development stage.
    - Export diversification is above what would have been predicted by its small population size.
  - Direct comparisons (SITC 1- and 4-digit):
    - San Marino is currently more diversified than Iceland and Andorra.
    - San Marino is less diversified than Malta, Luxembourg, Estonia, and Latvia.
    - At SITC 1-digit, San Marino is as diversified as Mauritius.
    - At SITC 4-digit, San Marino is more diversified than Cyprus and Switzerland.
  - Time trends:
    - At the 1-digit level, San Marino’s HHI indicates a slight decline in diversification since 2000.
    - At the 4-digit level, the decline in diversification is more pronounced, especially since 2016.

### Panel regression evidence: HHI vs GDP per capita and population (2000–2022)
- Data and method:
  - Annual data from 56 OECD and European countries during 1988−2023; panel regression with country-fixed effects of HHI on (i) per capita GDP, (ii) per capita GDP squared, and (iii) population size. All variables entered in natural logs.
- Main regression findings (Table 1; dependent variable Ln(HHI); sample: 2000-2022; # Countries = 56; Observations = 1788):
  - The relationship between HHI and per capita GDP is U-shaped (convex), confirmed across SITC disaggregation levels (1 to 5 digits):
    - Coefficient of Ln (GDP per capita) is always negative and highly statistically significant.
    - Coefficient of Ln (GDP per capita)^2 is always positive and highly statistically significant.
  - Implied elasticity ranges: results suggest a reduction of 1 percent in HHI for increases in per capita GDP ranging from 0.7 percent (SITC 1-digit) to 3.3 percent (SITC 5-digit).
  - Population effects:
    - For small economies and microstates there is a statistically significant negative relationship between HHI and population size.
    - In the wider country sample, the statistically significant negative relationship with population disappears in most specifications; the coefficient of population is not statistically significant in all but one specification.
    - An exception: at the SITC 5-digit level a statistically significant positive HHI-population association is found.
- Selected coefficient estimates from Table 1 (preserved exactly as reported):
  - Ln (GDP per capita): -0.73***, -0.84***, -1.84***, -1.64***, -2.33***, -2.04***, -1.94***, -1.93***, -3.32***, -2.45***.
  - Ln (GDP per capita^2): 0.04***, 0.04***, 0.10***, 0.08***, 0.13***, 0.11***, 0.13***, 0.13***, 0.18***, 0.13***.
  - Ln (Population) (selected reported values): -0.06, 0.12, 0.17, 0.00, 0.50***.
  - Constant (selected reported values): 2.13**, 3.58**, 6.23***, 3.47, 7.18***, 3.25, 3.33, 3.31, 10.59***, -1.11.
  - Note: * p-value < 0.1; ** p-value < 0.05; and *** p-value < 0.01.
- Interpretation for San Marino:
  - San Marino’s data are broadly clustered around the fitted regression curve for both SITC 1- and 4-digit levels, in a portion of the curve that surpassed the threshold level of GDP per capita where the relationship becomes positive.
  - Overall, San Marino is more diversified than its size would predict; for 1- and 4-digit disaggregation San Marino’s HHI is at or below the partial regression curve of HHI on population size.

### Entropy Index (EI): characteristics and San Marino’s position
- Definition and sensitivity:
  - EI captures the dispersion or spread in the distribution of export products, summing each export product’s market share multiplied by its natural log (and typically presented with a minus sign so higher EI means more diversified).
  - EI treats product shares more equally and is more sensitive to the presence of very small shares than the HHI.
- Cross-country comparisons and time patterns:
  - EI-based comparisons broadly confirm HHI-based results.
  - At SITC 1- and 4-digit levels, San Marino appears:
    - More diversified than Iceland and Andorra.
    - More concentrated than Luxembourg, Estonia, and Latvia.
    - At SITC 4-digit, more diversified than Cyprus and Malta, and at par with Switzerland.
  - EI also shows a decline in diversification for San Marino since the mid-2010s.
- Interpretation:
  - Similar overall pictures from HHI and EI are consistent with an evenly distributed export structure in San Marino: while some products dominate, San Marino does not display a very long tail of minor exports compared with peers.

### Export Concentration Ratio (CR-4): focused measure of top goods
- Definition and sensitivity:
  - CR-N sums the shares of the top N export products; CR-4 focuses on the top 4 export products at a given SITC level. It measures concentration at the top of the distribution and does not account for the remainder of the export basket.
- San Marino’s CR-4 findings:
  - At SITC 1-digit:
    - San Marino is currently less CR-diversified than most peers, except Andorra and, by a small margin, Iceland and Switzerland.
    - San Marino’s CR-4 has displayed a secular decline since the early 2000s at this aggregation level.
  - At SITC 4-digit:
    - The relative ranking aligns with HHI- and EI-based comparisons: San Marino is more diversified than Iceland, Andorra, Cyprus, and Switzerland by CR-4.
    - San Marino is becoming less diversified more recently at the 4-digit level as well.
- Interpretation:
  - The top 4 export products are more concentrated at an aggregated level in San Marino; reliance on a small number of key products reduces with more product disaggregation.

### Concluding remarks and policy implications
- Data and approach:
  - The analysis uses aggregate sectoral GVA from 13 economic sectors and three standard export diversification metrics applied to disaggregated export data constructed from mirror import data (San Marino values calculated using mirror import data from trading partners).
  - A strong Pearson Product Moment Correlation (PPMC) between aggregate sectoral GVA and export distributions justifies using export indicators as proxies for economic diversification.
- Sectoral structure:
  - Aggregate sectoral GVA points to concentration in manufacturing and wholesale and retail; these sectors are responsible for large shares of output and exports and, together with public administration, a sizeable share of employment.
  - Manufacturing and wholesale and retail shares on GVA are elevated and increasing; ICT is also increasing—these patterns could suggest rising economic concentration.
- Synthesis of index-specific evidence:
  - At aggregate sectoral levels San Marino appears concentrated (manufacturing, wholesale and retail).
  - At more disaggregated product levels San Marino emerges as a well-diversified economy:
    - HHI indicates a relatively diversified export structure compared to peer microstates.
    - EI suggests a balanced distribution relative to peers (no pronounced long tail of minor exports).
    - CR-4 shows a concentration in the top 4 goods at the aggregated 1-digit level, but this concentration disappears at the disaggregated 4-digit level.
- Policy-relevant implication:
  - While a few key products may constitute a larger share of exports in aggregated data, detailed disaggregation shows multiple products with meaningful shares. Policies should weigh the aggregate-sector concentration signals against the more nuanced picture from disaggregated export-product data when designing diversification or resilience strategies.

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

### 38.      The well-diversified economic structure of San Marino is consistent with the resilience

### The well-diversified economic structure of San Marino is consistent with the resilience of its economy amidst global trade, geopolitical, and commodity price shocks.

### Resilience and recent growth developments
- San Marino’s economy relied on a new growth model that diversified away from financial services after the Global Financial Crisis.
- The economy has maintained positive growth following the post-pandemic successive shocks.
- Sectoral dynamics driving resilience:
  - Manufacturing sector retained the market share gained through nearshoring.
  - Tourism sector flourished on post-Covid regional demand boom.
  - Total employment reached record highs.

### Export and economic diversification findings
- San Marino’s export and economic diversification are broadly in line with what would be predicted by both its size and level of development.
- Export diversification indicators for San Marino suggest that the country is more diversified than its size predicts.
- The paper confirms the inverted U-shaped relationship between development level and diversification found in earlier studies:
  - For poor countries, per capita GDP increases with more diversification, but after a certain threshold this relationship inverts and higher per capita GDP becomes associated with less diversification.
  - The analysis shows San Marino’s per capita GDP has surpassed this inflection point, suggesting that it may not benefit from further increases of GDP per capita associated with more diversification and that the recent slight decline in diversification observed since the mid-should not be a concern.

### Policy implications and recommendations
- Given San Marino is a well-diversified economy with a competitive manufacturing sector and emerging dynamic sectors like ICT, horizontal rather than sector-specific structural policies are recommended.
- Suggested horizontal reforms to strengthen existing productive sectors and boost economy-wide productivity:
  - Investing in both physical and digital infrastructure.
  - Increasing labor market flexibility.
  - Streamlining and modernizing the tax system.
  - Resolving banking-real estate distortions to facilitate real estate market functioning.
  - Enhancing public administration efficiency.
- Focus on higher value-added products within existing competitive sectors (manufacturing, ICT, professional services) rather than industrial policies to develop new sectors.
- Note: Salinas (2021) emphasizes the importance of horizontal policies and proximity to markets in explaining export diversification.

### Annex I — Export Diversification Indices (definitions preserved)
- Herfindahl-Hirschman Index (HHI):
  - Formula: 퐻퐻퐻퐻퐻퐻 = ∑푠푖2푁푖=1, where 푠푖 is the share of product 푖 in total exports and 푁 is the total number of products.
  - Interpretation: A lower HHI indicates a more diversified export base; a higher HHI indicates concentration on fewer products.
- Entropy Index:
  - Formula: 퐸퐸퐸퐸퐸퐸퐸퐸퐸퐸퐸퐸 = − ∑푠푖푁푖=1 . ln 푠푖, where 푠푖 is the share of product 푖 in total exports and 푁 is the total number of products.
  - Interpretation: A higher entropy value indicates a more diversified export structure.
- Export Concentration Ratio (CR):
  - Definition: Share of the top few products in total exports (e.g., CR-N measures the concentration of the top N export products).
  - Interpretation: A lower ratio indicates a more diversified export sector.

### Annex II — Additional empirical results (select regression outputs and sample information preserved)
- Panel regression: Entropy vs Country Size and Per Capita GDP, 2000-2022 (Table 1a, dependent variable: Entropy Index at various SITC digit levels)
  - Ln (GDP per capita): 0.47***, 0.60***, 1.41***, 1.25***, 1.99***, 1.53***, 3.11***, 2.37***, 3.27***, 3.40***
  - Ln (GDP per capita^2): -0.02***, -0.03***, -0.07***, -0.06***, -0.11***, -0.08***, -0.17***, -0.13***, -0.18***, -0.19***
  - Ln (Population): 0.07*, -0.09, -0.26***, -0.42***, 0.07
  - Constant: -0.89, -2.55**, -3.68***, -1.49, -4.94***, 1.19, -9.25***, 0.63, -10.56***, -12.31***
  - Observations1788 (repeated across columns)
  - # Countries56 (repeated across columns)
  - Note: * p-value < 0.1; ** p-value < 0.05; and *** p-value < 0.01.
- Panel regression: Export Concentration Ratio (Ln(CR)) vs Country Size and Per Capita GDP, 2000-2022 (Table 1b)
  - Ln (GDP per capita): -0.08*, -0.04, -1.23***, -0.95***, -1.70***, -1.38***, -1.46***, -1.32***, -1.31***, -0.48
  - Ln (GDP per capita^2): 0.00, 0.00, 0.07***, 0.05***, 0.10***, 0.08***, 0.09***, 0.09***, 0.07***, 0.03
  - Ln (Population): 0.02*, 0.16***, 0.19***, 0.08, 0.47***
  - Constant: 0.36, -0.21, 4.84***, 1.06, 6.23***, 1.90, 3.91***, 2.04, 3.72**, -7.27***
  - Observations1788 (repeated across columns)
  - # Countries56 (repeated across columns)
  - Note: * p-value < 0.1; ** p-value < 0.05; and *** p-value < 0.01.

*República of San Marino — International Monetary Fund*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2025/english/1smrea2025002-source-pdf.pdf_