## 1. Export Market Share and Growth of Estonia’s Top Export Markets

## Source details

**Canonical URL:** [1. Export Market Share and Growth of Estonia’s Top Export Markets](https://www.imf.org/-/media/files/publications/cr/2024/english/1estea2024002.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2024/english/1estea2024002.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2024/english/1estea2024002.pdf.json)

---

### Introduction and immediate drivers
- Estonia’s merchandise exports as a share of global exports have steadily declined for nine consecutive quarters, falling by 23 percent between 2021 Q3 and 2023 Q4.
- Immediate proximate causes identified:
  - Russia's war on Ukraine: extensive supply chain disruptions and increased input costs.
  - Substitution away from commodities previously supplied from Russia and Belarus to more expensive markets (timber and metals cited).
  - Euro appreciation against currencies of important trading partners (e.g., Sweden and Norway), diminishing competitiveness for Estonia’s medium- and low-tech exporters who are relatively price-sensitive.
  - Underperformance of key external markets, particularly the Scandinavian construction sector.
- Evidence that deterioration predates these shocks:
  - Productivity growth began faltering shortly after the global financial crisis and failed to keep pace with appreciation of the real exchange rate.
  - Relative costs in Estonia increased more than in the euro area, and recent price and cost adjustments exacerbated the trend.

### Decline in export market share — destination contributions
- Analysis focused on Estonia’s top ten export destinations (top ten based on export data as in 2023 Q3) plus residual EU and non-EU aggregates; by construction contributions add to the total observed decline.
- Aggregate statistic:
  - Total export share decline (Estonia, 2021 Q3-2023 Q4): -23%
- Largest negative contributors to the export share decline (2021 Q3–2023 Q4):
  - United States (largest negative contribution despite being the 10th largest destination of Estonia’s exports in 2023).
  - Netherlands.
  - Latvia.
- Sweden (absorbing 9 percent of Estonia’s exports in 2023) ranks fourth in contributions to the decline.
- Role of Russia was limited (Russia is not a key export market for Estonia); Russia’s negative contributions reflect sanctions-related effects in some decompositions.

### Methodology — Constant Share Analysis (CSA)
- CSA decomposes changes in a country’s share of world exports into:
  - Intensive margin (competitiveness effect): change in the country’s penetration of partner markets (change in the country’s share of partner’s imports), holding partner size constant.
  - Extensive margin (composition effect or foreign demand): effect of growth in the size of export partners, holding relative penetration constant.
  - Interaction term: the product of changes in penetration and partner size.
- Formal identity (one product, one partner case) preserved from source notation:
  - θ_r = θ_rW × δ_W and ∆θ_r = ∆θ_rW δ_W^0 + ∆δ_W θ_rW^0 + ∆θ_rW ∆δ_W
- Limitations noted:
  - Assumes market structure unchanged; ignores technological change, consumer preferences, regulatory changes, competitor actions, and exchange rate effects.
  - Assumes product homogeneity and does not provide causal identification.

### Key analytical findings and interpretation
- Majority of Estonia’s export share decline attributable to the intensive margin (shrinking share of Estonia’s exports in main destination markets) — interpreted as weakening external competitiveness and declining relative productivity.
- The decline is not solely cyclical due to global demand composition; structural factors (relative productivity and competitiveness) play an important role.
- CSA useful to quantify geographic and product concentration effects and competitor penetration when direct industry-level relative productivity comparisons are unavailable.

### CSA decomposition — Estonia (2021 Q3–2023 Q4) and time variation
- Main empirical result (2021 Q3–2023 Q4):
  - Decline largely driven by intensive margin (competitiveness).
  - Small fraction of share declines in Finland, Latvia, and Norway due to composition effects (shrinking share of world trade claimed by those countries).
  - Russia: negative contributions from both composition and competitiveness effects, likely reflecting EU trade sanctions.
- Time-interval variation:
  - 2010 Q1-2011 Q1: increase in Estonia’s export market share mainly attributable to a strong intensive-margin contribution.
  - 2011 Q2-2021 Q2: broadly flat export share driven by a negative composition effect offsetting positive competitiveness contributions.
  - Attribution becomes less precise over longer intervals.

### Comparison across the Baltics (post-Covid aggregates)
- Aggregate export share changes (post-Covid, 2021 Q3–2023 Q4):
  - Estonia: -23%
  - Latvia: -6%
  - Lithuania: -7%
- Decomposition contrasts:
  - Estonia: majority of decline attributed to the intensive margin (competitiveness).
  - Latvia: large contribution from the interaction term (country-level interaction mainly associated with Russia).
  - Lithuania: fall mainly reflected a large contribution from the extensive margin (shrinking foreign demand).
- Period-specific observations:
  - 2010 Q1-2011 Q1: very large competitiveness gains underpinned strong export market dynamics, particularly in Estonia and Lithuania.
  - 2011 Q2-2021 Q2: strong export market dynamics in Latvia and Lithuania driven by competitiveness gains; Estonia’s export share remained flat.

### Sector contributions to export performance (2-digit HS commodity categories)
- Estonia (2021 Q3-2023 Q4):
  - Losses of export shares mainly due to the intensive margin.
  - Concentration of losses in Mineral products, Wood and wood products, and Machinery/electrical products.
- Latvia:
  - Mixed picture: losses in some sectors partly offset by gains in others.
- Lithuania:
  - Small decrease in export share mainly owes to the intensive margin.
- Product-level totals reiterated:
  - Estonia product-level total export share decline: -23%
  - Latvia product-level total export share decline: -6%
  - Lithuania product-level total export share decline: -6%

### Services export share and Export Performance Index (EPI)
- Structural-transformation consideration:
  - Shift from goods to services could explain merchandise trade decline given services’ rising share in gross value added (notably ICT) and net external balance in services.
- Data constraints:
  - CSA decomposition for services export share cannot be performed due to lack of data.
- Export Performance Index evidence:
  - EPI shows Estonia’s services exports have increased by less than foreign demand since 2021, indicating some loss of competitiveness in services.
  - Services account for only about a third of total exports; large gains in services would be required to offset sharp falls in goods exports.
- EPI definition and presentation:
  - EPI defined as ratio between country i’s export volumes and its export demand (weighted average of import volumes of trading partners using trade weights reflecting destination relevance).
  - EPI series shown with 2010=100; staff estimate used for 2023 data point.

### Conclusions on competitiveness and policy implications
- Overarching finding:
  - Estonia's falling export market share signals a persistent challenge in external competitiveness.
  - Immediate shock from Russia's war in Ukraine contributed, but faltering productivity and progressive softening of external competitiveness predate recent shocks and have been evident since the global financial crisis.
- Specific conclusions:
  - Unlike Latvia and Lithuania, majority of Estonia's export market share decline can be attributed to the intensive margin (diminishing share in key destination markets), signaling declining external competitiveness and relative productivity.
  - Services growth lagged foreign demand since 2021.
- High-level policy implications:
  - Address erosion of external competitiveness through structural reforms to enhance productivity.
  - Remove impediments to structural transformation toward more technologically intensive and higher value-added products and services.
  - Ensure real wage growth remains closely aligned with productivity growth.
  - Restoring external competitiveness is necessary to ensure continued convergence towards income levels of EU most advanced economies and Nordic neighbors.

### Annex / additional notes (selected)
- Annex I: country contributions to decline in aggregate export market shares for Latvia and Lithuania (2021 Q3-2023 Q4):
  - Latvia: total export share decline: -6%; losses mainly reflected a strongly negative interaction term in Russia (likely reflecting EU sanctions effects).
  - Lithuania: total export share decline: -7%; small overall decrease mainly driven by Russia.
- References and methodological sources cited in the chapter include Di Mauro et al. (2005), Gaulier and Zignago (2010), Gilbert (2017), Mandel (2012), Richardson (1971).

*Prepared by Gianluigi Ferrucci and Sadhna Naik; May 23, 2024. International Monetary Fund.*

### 1. Export Market Share and Growth of Estonia’s Top Export Markets ___________________ 5

### 1. Export Market Share and Growth of Estonia’s Top Export Markets ___________________ 5

### Introduction
- Estonia’s export market share has fallen sharply in recent quarters, both in absolute terms and relative to comparable economies.
- Having remained remarkably stable throughout the COVID-19 pandemic, Estonia’s merchandise exports as a share of global exports have steadily declined for nine consecutive quarters, falling by 23 percent between 2021 Q3 and 2023 Q4.
- Immediate proximate causes identified:
  - Russia's war on Ukraine: extensive supply chain disruptions and increased input costs.
  - Substitution away from commodities previously supplied from Russia and Belarus to more expensive markets (timber and metals cited).
  - Euro appreciation against currencies of important trading partners (e.g., Sweden and Norway), diminishing competitiveness for Estonia’s medium- and low-tech exporters who are relatively price-sensitive.
  - Underperformance of key external markets, particularly the Scandinavian construction sector.
- Evidence that deterioration predates these shocks:
  - Productivity growth began faltering shortly after the global financial crisis and failed to keep pace with appreciation of the real exchange rate.
  - Relative costs in Estonia increased more than in the euro area, and recent price and cost adjustments exacerbated the trend.

### Decline in Estonia’s Export Market Share — Destination Contributions
- Analysis focuses on Estonia’s top ten export destinations (top ten based on export data as in 2023 Q3) plus a residual category split into EU and non-EU aggregates.
- By construction, contributions across all countries and country groups must add up to the 23 percentage points decline observed in Estonia’s export market share between 2021 Q3 and 2023 Q4.
- Largest negative contributors to the export share decline (2021 Q3–2023 Q4):
  - United States (largest negative contribution despite being the 10th largest destination of Estonia’s exports in 2023).
  - Netherlands.
  - Latvia.
- Sweden, although a large destination (absorbing 9 percent of Estonia’s exports in 2023), ranks fourth in contributions to the decline.
- Role of Russia was limited (Russia is not a key export market for Estonia).

### Methodology — Constant Share Analysis (CSA) (intuitive exposition)
- CSA decomposes changes in a country’s share of world exports into:
  - Intensive margin (competitiveness effect): change in the country’s penetration of partner markets (change in the country’s share of partner’s imports), holding partner size constant. Interpreted as attributable to competitiveness gains/losses.
  - Extensive margin (composition effect): effect of growth in the size of export partners, holding relative penetration constant. Measured by weighing the change in the size of an export destination market in world trade by the average share of the country’s exports to that destination.
  - Interaction term: the product of changes in penetration and partner size.
- Formal notation (special case: one product, one partner):
  - Let X_r be total exports of economy r and X_W total world exports. Country’s share: X_r / X_W.
  - Let X_Wp be world exports to country p. Then country r’s export share = (X_r / X_Wp) × (X_Wp / X_W).
  - Define θ_r = X_r / X_W, θ_rW = X_r / X_Wp, δ_W = X_Wp / X_W.
  - Then θ_r = θ_rW × δ_W.
  - With ∆ denoting change between two periods:
    - ∆θ_r = ∆θ_rW δ_W^0 + ∆δ_W θ_rW^0 + ∆θ_rW ∆δ_W
    - Shares δ and θ are evaluated at their initial values in the decomposition.

### Key analytical findings and interpretation
- A significant portion of Estonia’s export share decline is attributable to the intensive margin (shrinking share of Estonia’s exports in main destination markets), indicating weakening external competitiveness and declining relative productivity—this pattern differs from Latvia and Lithuania.
- The decline is not solely a cyclical phenomenon driven by shifts in global demand composition; structural factors (relative productivity and competitiveness) play an important role.
- The CSA provides a means to quantify the extent to which geographic and product concentration, and changes in competitor penetration, explain the export share loss when direct industry-level relative productivity comparisons are not available.

*Prepared by Gianluigi Ferrucci and Sadhna Naik; May 23, 2024. International Monetary Fund.*

### 10.      Gilbert (2017) provides an intuitive geometric exposition of the CSA breakdown.

### 1estea2024002 - 10.      Gilbert (2017) provides an intuitive geometric exposition of the CSA breakdown.

### CSA methodology and decomposition
- Gilbert (2017) provides an intuitive geometric exposition of the CSA breakdown. The change in the export share is represented as an area that can be decomposed into three components: the intensive margin (competitiveness effect), the extensive margin (composition effect or foreign demand), and an interaction term.
- Algebraic identity presented (using source notation):
  - Let zz = xxxx, then: ∆zz = xxxx − xx0/xx0.
  - Adding and subtracting xxxx0: ∆zz = xx∆xx + xx0∆xx.
  - Further rearrangement yields: ∆zz = xx0∆xx + xx0∆xx + ∆xx∆xx (as presented in the source).
- Limitations of CSA noted:
  - Assumes market structure remains unchanged over the analyzed period.
  - Ignores technological advancements, changes in consumer preferences, regulatory changes, exchange rate changes, competitor actions.
  - Assumes product homogeneity; does not provide causal identification of underlying drivers.
  - For a comprehensive discussion of limitations, Richardson (1971) is cited.

### CSA decomposition: Estonia — key findings
- Objective: distinguish changes in Estonia’s export market share attributable to intensive margin (competitiveness/productivity) versus composition effects (size of destination markets).
- Main empirical result for 2021 Q3 to 2023 Q4:
  - The decline in Estonia’s export share has been largely driven by the intensive margin (shrinking market share in destination countries, reflecting falling competitiveness).
  - Only a small fraction of share declines in a handful of countries (notably Finland, Latvia, and Norway) was due to composition effects (shrinking share of world trade claimed by those countries).
  - For Russia, negative contributions from both composition and competitiveness effects likely reflect, at least in part, the effects of the EU trade sanctions in response to Russia's invasion of Ukraine.
- Time-interval variation:
  - 2010 Q1-2011 Q1: increase in Estonia’s export market share was mainly attributable to a strong contribution from the intensive margin.
  - 2011 Q2-2021 Q2: broadly flat export share driven by a negative composition effect offsetting positive competitiveness contributions.
  - Attribution becomes less precise over longer time intervals.

- Exact aggregate statistics cited:
  - Total export share decline (Estonia, 2021Q3-2023Q4): -23%

### Comparison across the Baltics (Latvia, Lithuania, Estonia)
- Post-Covid period aggregate export share changes:
  - Estonia: -23%
  - Latvia: -6%
  - Lithuania: -7%
- Decomposition differences:
  - Estonia: majority of decline attributed to the intensive margin (competitiveness).
  - Latvia: large contribution from the interaction term (country-level interaction mainly associated with Russia).
  - Lithuania: fall mainly reflected a large contribution from the extensive margin (shrinking foreign demand).
- Period-specific observations:
  - 2010 Q1-2011 Q1: very large competitiveness gains underpinned strong export market dynamics, particularly in Estonia and Lithuania.
  - 2011 Q2-2021 Q2: strong export market dynamics in Latvia and Lithuania driven by competitiveness gains; Estonia’s export share remained flat.

### Sector contributions to export performance
- CSA applied to 2-digit HS commodity categories (Eurostat data; top 10 commodity exports as in 2023Q3).
- Estonia (2021 Q3-2023 Q4):
  - Losses of export shares mainly due to the intensive margin.
  - Concentration of losses in Mineral products, Wood and wood products, and Machinery/electrical products.
- Latvia:
  - Mixed picture: losses in some sectors partly offset by gains in others.
- Lithuania:
  - Small decrease in export share mainly owes to the intensive margin.
- Total export share declines by product (Estonia): total export share decline: -23%
- Latvia product-level total export share decline: -6%
- Lithuania product-level total export share decline: -6%

### Services export share and Export Performance Index (EPI)
- Structural-transformation argument: shift from goods to services might explain merchandise trade decline given services’ rising share in gross value added (notably ICT) and net external balance in services.
- Data limitations: CSA decomposition for services export share cannot be performed due to lack of data.
- Export Performance Index (EPI) evidence:
  - EPI shows Estonia’s services exports have increased by less than foreign demand since 2021, indicating some loss of competitiveness in services.
  - Services still account for only about a third of total exports; therefore, large gains in services would be required to offset sharp falls in goods exports.
- EPI and structural notes:
  - Export Performance Index defined as ratio between country i’s export volumes and its export demand (weighted average of import volumes of trading partners using trade weights that reflect destination relevance).
  - EPI series shown with 2010=100; staff estimate used for 2023 data point.
- Share of goods and services in total exports (percent) and Current Account Balance (Percent of GDP) series are presented in the source figures.

### Conclusions and policy implications
- Overarching finding:
  - Estonia's falling export market share indicates a persistent challenge in external competitiveness.
  - Immediate cause includes the external shock from Russia's war in Ukraine, but faltering productivity and progressive softening of external competitiveness predate recent shocks and have been evident since the global financial crisis.
- Specific conclusions:
  - Unlike Latvia and Lithuania, the majority of Estonia's export market share decline can be attributed to the intensive margin (diminishing share in key destination markets), signaling declining external competitiveness and relative productivity.
  - Services growth also lagged foreign demand since 2021, indicating challenges in services exports.
- High-level policy implications:
  - Address erosion of external competitiveness through structural reforms to enhance productivity.
  - Remove impediments to structural transformation toward more technologically intensive and higher value-added products and services.
  - Ensure real wage growth remains closely aligned with productivity growth.
  - Restoring external competitiveness is necessary to ensure continued convergence towards income levels of EU most advanced economies and Nordic neighbors.

### Annex / Additional results (selected)
- Annex I: country contributions to decline in aggregate export market shares for Latvia and Lithuania (2021 Q3-2023 Q4).
  - Latvia: total export share decline: -6%; losses mainly reflected a strongly negative interaction term in Russia (likely reflecting EU sanctions effects).
  - Lithuania: total export share decline: -7%; small overall decrease mainly driven by Russia.
- References cited in the source for methods and context include Di Mauro et al. (2005), Gaulier and Zignago (2010), Gilbert (2017), Mandel (2012), Richardson (1971).

*Source: Selected excerpts from the IMF chapter on Estonia (CSA decomposition, sector and services analysis, conclusions, and annex figures).*

### 4. The same methodology is extended to the other two Baltic countries, aiming at

### 1estea2024002 - 4. The same methodology is extended to the other two Baltic countries, aiming at

### Methodology
- Step 1: Estimate TFP from a Cobb-Douglas production function using quarterly data:
  - Equation in logs: a_t = x_t − (1−α_t) k_t − α_t l_t
  - Variables:
    - x_t and k_t measured in millions of constant 2015 euros.
    - l_t measured in thousands of hours-worked per quarter.
    - Effective capital stock k_t constructed from annual capital stock (AMECO), quarterly investment flows, depreciation estimates, and multiplied by industrial capacity utilization.
    - Labor input constructed as employees (e_t) × hours per employee (h_t).
    - Labor share α_t: a smooth, time-varying trend of the ratio of compensation of employees to GDP (time-varying since ~2015 for Latvia and Lithuania; relatively stable and mean-reverting for Estonia).1
- Step 2: Isolate structural TFP and estimate potential GDP using a multivariate Kalman filter and state-space decomposition:
  - TFP series from Step 1 contains cyclical and structural components.
  - Potential GDP x̂_t estimated from state-space decomposition of trend and cycle using monthly confidence indices (consumer, industry, construction, retail), unemployment rate, and industrial capacity utilization as signal variables.
  - Trend TFP â_t obtained by applying the production function to HP-filter trends for effective capital k̂_t, employment ē_t, and hours-worked ĥ_t.
- Step 3: Test and estimate cointegration between TFP and REER to assess the Balassa-Samuelson hypothesis:
  - Augmented Dickey-Fuller tests used for integration with constant and linear trend; lags via Schwartz information criteria.
  - Johansen cointegration tests confirm at least one cointegrating relationship between ln(TFP) and ln(REER) for Estonia (and confirmed for all three Baltic countries).
  - Dynamic Least Squares (DOLS) used to estimate cointegrating equation (ln(REER) dependent).  
  - Cointegration sample and estimation details for Estonia:
    - Johansen sample (adjusted): 1995Q4 2023Q2; Number of observations: 111; Endogenous variables: ln (TFP) ln (REER).
    - DOLS sample (adjusted): 1996Q4 2023Q2; Included observations: 107; Cointegrating equation deterministic factors: C @TREND; Automatic leads and lags: lead=0 and lag=6 based on SIC criterion.
  - Example cointegration regression snippet:
    - Ln (TFP) coefficient: 0.474694; Std. Error: 0.0995334; t-Statistic: 4.769219; Prob.: 0.0000
    - C: 117.86762; Std. Error: 9.786223; t-Statistic: 23.957117; Prob.: 0.0001
    - @TREND: 0.318322; Std. Error: 0.016930; t-Statistic: 18.802480; Prob.: 0.0000
    - R-squared: 0.963722; Adjusted R-squared: 0.960356; S.E. of regression: 2.478203; Long-run variance: 18.70264.

1 Footnote in source: For Estonia, α_t is relatively stable and mean-reverting but it has been increasing markedly in both Latvia and Lithuania since around 2015. For that reason, in the estimation of total factor productivity, instead of a fixed calibrated value for α (e.g., at its historical average or last observed value) a smooth, time-varying labor share was used for the three countries.

### Key Results — Potential GDP and Drivers
- Potential GDP and scarring:
  - Simple extrapolations (Figure 3) suggest significant output losses following the GFC and, to a lesser extent, the pandemic.
  - Potential GDP growth fell from above 5 percent pre-GFC to around 2 percent post-GFC.
  - Actual growth dropped even lower after 2020 and well below potential growth, indicating a significant cyclical component in the current downturn.
- Quantified changes in growth rates (Estonia): Changes between 1995-2008 and 2020-2023Q2 (percentage points):
  - GDP: Total −4.3; Trend −3.1; Cyclical −1.2
  - TFP: Total −5.2; Trend −2.7; Cyclical −2.5
  - K (capital): Total −1.1; Trend −2.3; Cyclical 1.1
  - L (labor): Total 1.3; Trend 1.3; Cyclical 0.1
  - Δα (change in labor share): Total 0.7; Trend 0.7; Cyclical 0.0
- Decomposition highlights:
  - The fall in TFP growth accounts for most of the decline in GDP growth: the decline in average TFP growth surpasses that of actual GDP growth and explains almost 90 percent of the fall in estimated potential GDP growth (Table 3 and narrative).
  - Capital has been the major driver of GDP growth historically; capital’s contribution remains larger than labor’s even recently.
  - Labor contribution shifted from a drag pre-GFC to supporting growth in subsequent periods, but its contribution is still less significant than capital despite migration flows after Russia’s war on Ukraine.
  - Capital accumulation and capacity utilization have cushioned the decline in GDP growth after 2020, although capacity utilization remains on a downward trend.
  - Demographics, labor participation, and hours-worked have contributed positively to GDP growth recently, while the structural contribution of the unemployment rate has turned negative, suggesting an increase in the natural rate of unemployment.

### TFP and REER Interaction — Competitiveness Implications
- Cointegration and Balassa-Samuelson:
  - Cointegration tests are consistent with the Balassa-Samuelson hypothesis: long-run relationship between TFP and REER found for all three Baltic countries.
  - The fitted cointegrating relationship is used to construct a TFP-based REER; gaps between actual and TFP-based REER measure deviations from the Balassa-Samuelson implied equilibrium.
  - Interpretation: negative gaps = price-competitiveness advantage; positive gaps = price-competitiveness disadvantage.
- Estonia’s competitiveness dynamics:
  - Pre-GFC: fast TFP growth produced negative REER gaps (REER undervalued relative to TFP-based REER), providing a price-competitiveness edge.
  - Post-2008: negative gaps became smaller and shorter-lived; positive gaps became larger and more frequent, indicating erosion of competitiveness.
  - Since late 2021: a significant divergence between actual and TFP-based REER started as TFP deceleration turned into decline and compounded with real exchange rate appreciation. This overvalued exchange rate likely reduced Estonia’s ability to absorb shocks, leading to a decline in net exports and economic activity.

### Regional Comparison Across the Baltics
- Common pattern:
  - Capital accumulation drove the Baltics’ convergence pre-GFC (capital deepening and fast income convergence).
- Divergence after the GFC:
  - Latvia and Lithuania: TFP has become an increasing driver of growth, especially Lithuania which shows a structural TFP contribution to expansion.
  - Estonia: TFP declined and eventually turned negative, contributing to underperformance after 2021.
- Cyclical vs structural components:
  - Latvia and Estonia: cyclical components (especially TFP cyclical swings) play an important role.
  - Lithuania: expansion largely driven by structural TFP and, to a lesser extent, capital.
- REER gaps across Baltics:
  - Pre-GFC: Estonia and Latvia had consistently negative REER gaps (competitiveness advantage); Lithuania only turned negative in 2005–2007 and to a smaller extent.
  - Post-GFC: Estonia’s competitiveness disadvantage started earlier and became stronger than in Latvia and Lithuania; REER gap larger for Estonia consistent with underperformance.

### Conclusions
- Potential GDP growth in Estonia has fallen since the GFC, largely due to a steady decline in TFP growth.
- While some decline in potential GDP growth accompanies income convergence (as capital accumulation decelerates), the largest contributor to Estonia’s reduction in potential GDP growth has been the decline in TFP growth.
- The drop in TFP growth in Estonia has been significantly more pronounced than in Latvia and, especially, Lithuania (which appears to experience an acceleration in TFP growth in recent years).

*Sources: Statistics Estonia; Eurostat; Haver Analytics; and IMF staff calculations.*

### 22.      Differently from Latvia and Lithuania, the level of TFP has declined in Estonia since

### 22.      Differently from Latvia and Lithuania, the level of TFP has declined in Estonia since

### TFP decline and competitiveness
- Estonia: TFP growth has declined since the GFC, and has become negative more recently.
- The decline in TFP growth has a structural component, likely associated with the scarring effects of recent shocks.
- Implications for competitiveness:
  - Pre-GFC: fast TFP growth underpinned Estonia’s competitive advantage despite real exchange rate appreciation.
  - Post-GFC: decelerating TFP growth eroded Estonia’s competitive advantage.
  - More recently: significant real exchange rate appreciation compounded declining TFP, turning into a competitive disadvantage and increasing vulnerability to recent shocks.
  - The positive wedge between the actual and TFP-based REER in Estonia started earlier, evolved faster, and became wider than in Latvia and Lithuania, reducing Estonia’s external competitiveness by a much larger factor.

### Role of allocative efficiency in productivity growth (summary findings)
- Resource misallocation has dragged down total factor productivity (TFP) growth in all three Baltic economies.
- In Estonia:
  - Allocative efficiency improved after the global financial crisis (GFC), but the recovery was limited and short-lived.
  - Productivity loss due to resource misallocation is more pronounced for services than for goods sectors.
  - Some industries—such as mining and real estate—saw allocative inefficiency worsen more significantly.
- Quantitative rule-of-thumb from the methodology:
  - "For each unit decline in allocative efficiency, there will be a one-percentage point decline in TFP growth."

### Data, sample periods, and methodology (key points preserved)
- Data and sources:
  - Estonia and Latvia: Orbis dataset.
  - Lithuania: administrative data from Statistics Lithuania.
  - Sample periods noted in figures: Estonia (1997-2020), Lithuania (2000-2020), and Latvia (2010-2020).
- Methodological framework:
  - Follows Hsieh and Klenow (2009) and IMF World Economic Outlook (April 2024).
  - Assumptions include:
    1) A Cobb-Douglas production function at the firm level with 훼훼푐푐푐푐 as country-sector specific capital share and firm-level variables Y, A, K, L (subscripts c, s, i, t).
    2) Aggregation with constant elasticity of substitution where 휎휎푐 represents the elasticity of substitution.
    3) Distortions on output, capital, and labor markets with effective cost increases 휏휏퐾 and 휏휏퐿 and an output tax 휏휏푌; 휏휏푐푐푐푐푐푐푡푡 defined as function of these distortions.
    4) Firms maximize profits under monopolistic competition, equating marginal revenue products to marginal costs (Equations (4) and (5)); sector-level costs r and w appear.
    5) In the ideal no-distortions case (휏휏 parameters equal 1), sector-level TFP aggregates firm-level technology A; with distortions TFP is lower and the wedge is AE (Equation (10)).
  - Aggregation and decomposition:
    - Aggregate allocative efficiency up to the sector level; decompose country-level TFP growth into innovation and allocative efficiency components (Equation (11)).

### Decomposition results and time patterns
- Overall: Allocative inefficiency has negatively contributed to TFP growth in Estonia, Latvia, and Lithuania over the sample period.
- Time variation:
  - Estonia: allocative efficiency contribution was largely negative before the GFC, turned positive during the post-crisis recovery period, but worsened again in recent years.
  - The post-GFC improvement in allocative efficiency in Estonia may have been driven by market corrections, reassessment of risk, structural reforms, business efficiency efforts, market discipline, and policy interventions.
- Within-sector vs across-sector:
  - In Estonia and many advanced economies, within-sector allocative efficiency is more important for productivity growth than reallocation across sectors.
  - The IMF Spring 2024 World Economic Outlook finds a significant negative contribution of within-sector allocative efficiency to annual TFP growth for most advanced euro-area economies, including Estonia; by contrast, within-sector allocative efficiency in the United States contributed positively to TFP growth.

### Sectoral patterns and dispersion indicators
- Services vs goods:
  - Productivity loss due to allocative inefficiency is estimated to be about 8 percent of TFP for the Baltic region on average, with a larger difference for Latvia and a smaller one for Lithuania.
  - Services sectors typically show more market frictions, product differentiation, market power, regulatory hurdles, licensing requirements, and entry barriers—factors that can prolong inefficiency.
  - Skill shortages in services (e.g., ICT and professional services) in Estonia may have resulted in allocative inefficiency.
- Specific sectors in Estonia:
  - Mining, electricity, and real estate experienced worsening resource misallocation in the years leading up to the GFC.
  - In recent years, allocative efficiency has deteriorated significantly for the mining sector; the analysis notes the need for further sector-specific investigation beyond this work.
- Dispersion metrics:
  - The variance across firms in the marginal revenue product of capital has increased steadily in Estonia (Figure 6).
  - The variance of TFPR (dispersion in productivity across firms) has increased over time (Figure 7).
  - Increasing dispersion is most prominent in mining and real estate (Figure 8), consistent with widening allocative inefficiency.

### Structural policy implications and recommendations
- Structural reforms can improve allocative efficiency, bolster productivity growth, and help restore competitiveness for Estonia.
- Types of reforms associated with better allocative efficiency (IMF evidence cited):
  - Less regulation in product markets.
  - Greater financial market liberalization.
  - More liberal labor markets (room for improvement noted for Estonia on labor market liberalization).
- Specific policy considerations:
  - Liberalize product markets where regulation constrains reallocation and competition.
  - Promote financial market liberalization to reduce distortions in capital allocation.
  - Increase labor market flexibility while managing protections—policies protecting jobs in downturns may hamper efficient labor allocation (example: job retention schemes during the pandemic may have led to productivity losses).
  - Undertake targeted analysis of mining and real estate sectors to identify frictions and tailor reforms.
- Cross-country benchmarking:
  - Indicators from the IMF Structural Reforms Dataset (latest period 2014) place Estonia favorably on product market regulation and financial liberalization relative to many advanced and emerging economies, but show scope for improvement on labor market liberalization.

*Prepared by Bingjie Hu and Can Ugur; sources: Orbis, EUKLEMS, Statistics Lithuania, IMF staff calculations, IMF World Economic Outlook (April 2024).*

### References

### References

### Citations
- Blanchard, Olivier J., Mark Griffiths, and Bertrand Gruss. "Boom, bust, recovery: Forensics of the Latvia crisis." Brookings Papers on Economic Activity 2013, no. 2 (2013): 325-388.
- Decker, Ryan A., John Haltiwanger, Ron S. Jarmin, and Javier Miranda. "Declining dynamism, allocative efficiency, and the productivity slowdown." American Economic Review 107, no. 5 (2017): 322-326.
- Hsieh, Chang-Tai, and Peter J. Klenow. "Misallocation and manufacturing TFP in China and India." The Quarterly Journal of Economics 124, no. 4 (2009): 1403-1448.
- International Monetary Fund, April 2024 “World Economic Outlook”. Available at https://www.imf.org/en/Publications/WEO/Issues/2024/04/16/world-economic-outlook-april-2024
- Meriküll, Jaanika and Alari Paulus, “Were jobs saved at the cost of productivity in the COVID-19 crisis?” European Economic Review, Volume 161, 2024, available at https://doi.org/10.1016/j.euroecorev.2023.104618

### A. Introduction — Key points
- Emerging needs to strengthen national security and accelerate the energy transition add to long-standing ageing-related pressures.
- Growing tension between retaining a historically competitive low tax environment and moving towards broader provision of public services and a stronger social safety net may lead to further fiscal deterioration, if left unaddressed.
- New spending needs stem from goals to strengthen national security and accelerate the green and energy transition, while ageing-related areas such as pensions and healthcare continue to exert pressure.
- If some new spending is permanent or contains a strong inertial component while new revenue measures follow with a lag, unfavorable dynamics for public finance and inefficient policy outcomes (fast increase in public debt and interest payments) may ensue.
- The paper performs a high-level benchmarking exercise comparing spending categories—by economic concept and by function—and revenue items across comparator groups: other Baltic neighbors (Latvia and Lithuania), Nordic countries (Denmark, Finland, Norway, and Sweden), and the Euro Area (EA20), to identify options for expenditure-based consolidation and revenue mobilization.

### B. Expenditures — Findings and statistics
- Estonia’s government expenditures as share of GDP:
  - Were the lowest among comparator groups and on a declining trend until 2007.
  - Increased by almost 3 pp between 2011 and 2023.
  - Over 2017-2023, Estonia’s government spending as share of GDP was on average only higher than that of Ireland, Lithuania, Romania, and Malta, and just below that of Latvia.
- Pandemic effects:
  - In the five years prior to the pandemic, Estonia’s general government spending averaged less than 38 percent of GDP, about 6.5 and 11 percentage points less than the average of the Euro Area (EA20) and the Nordic countries respectively.
  - In response to the pandemic, spending went up by about 5 pp of GDP, largely reversed in 2021-2022, but spending was on the rise again in 2023.
  - Post-pandemic decline in government spending as a share of GDP: Nordics fell by 8.3 percentage points; EA20 fell by 4.1 percentage points.
  - By 2023, the gap between Estonia and the Nordics was reduced to 8.5 percentage points of GDP, while remaining about the same compared with the EA20.
- Composition of spending (by economic transaction):
  - Estonia underspends peers on interests, subsidies, and social benefits (three categories driving relative underspending).
  - Estonia outspends the EA20 on government consumption (compensation of employees and purchase of goods and services) and public investment (capital expenditures), though it still lags the Nordics (except on capital expenditures).
- Trends 1995–2023 by transaction:
  - Rising trends for compensation of employees, capital expenditures, and social benefits since the post-GFC period.
  - Purchase of goods and services increased in response to the GFC but followed a more stable path afterward.
  - Interest expenditures declined given low public debt, with the trend reversing recently due to rising debt and higher interest rates.
- Contribution to the increase in government spending since 2011:
  - Compensation of employees and social benefits explained about 90 percent of the change.
  - Capital expenditures contributed with 3 pp.
  - In contrast, Nordics and EA20 reduced total expenditures as share of GDP in almost all categories, especially compensation of employees, social benefits, and interests.
- Structural differences and gaps:
  - Except for capital expenditures, Estonia still underspends both the EA20 and the Nordics in all categories of spending by economic transaction.
  - The gap in social benefits is about 3 pp of GDP (EA20) and 4 pp of GDP (Nordics).
- Competing pressures summarized:
  - (i) Estonia already outspends peers such as the EA20 on public investment and operational costs to provide public services (compensation of employees and purchase of goods and services).
  - (ii) Estonia will experience an increase in interest payments—one category where Estonia clearly underspends peers—which will divert resources to debt service.
  - (iii) Estonia may face demand for broader social safety nets and better-quality public services.
- Expenditure by function (2001–2022):
  - Estonia underspends EA20 and Nordic peers on health and social protection.
  - Estonia underspends the EA20 on environment protection.
  - Future demand for health, social protection, and environment protection likely to increase due to population ageing and green/energy transition commitments.
- Areas where savings appear unlikely:
  - Estonia outspends peers in defense and education, and the Nordics on R&D.
  - Spending in defense, education, R&D likely to remain given geopolitical tensions and the need to support productivity and economic transformation.
  - Ongoing decline of spending in education, R&D, and environment protection reduces their suitability to accommodate higher spending elsewhere.

### C. Revenues — Findings and statistics
- General government revenue as share of GDP:
  - Over 2017-2023, Estonia’s general government revenue has been on average significantly lower than that of the Nordics and the EA20, albeit slightly higher than that of the other Baltic countries.
  - Following a decreasing trend through 2005 with a trough of 34 percent of GDP, average government revenue in Estonia permanently increased by about 3 pp of GDP after the GFC, remaining broadly stable at about 38 percent of GDP since then.
  - As of 2023Q2, the Estonian government mobilized 39.3 percent of GDP, still about 13 pp and 4 pp of GDP less than the Nordic countries and the EA20 respectively.
- Counterfactual revenue assessment:
  - A time- and country-fixed effect panel regression of total government revenues as share of GDP on real per capita income over 2000-2022 suggests actual revenue collection is on average 2.7 pp of GDP below what would be expected given Estonia’s per capita income (except for 2009-2010).
  - This implies space for additional revenue mobilization alongside spending consolidation.
- Tax revenue composition and trends (1995–2022):
  - Tax revenues have generally increased since 2011, except for property and excise duties.
  - Between 2011 and 2023:
    - Personal income tax (PIT) increased by about 1 percentage point of GDP.
    - Value-added tax (VAT) increased by about 1 percentage point of GDP.
    - Corporate income tax (CIT) collection increased by about 0.6 pp of GDP.
    - Property tax revenues, already historically low, have declined further.
    - Excise duties generated less 1.5 pp of GDP in revenues since 2011.
  - The observed increase in tax revenues (2 percentage points of GDP) fell short of the increase in spending (+ 2.8 pp) over the same period.

*Prepared by Carlos de Resende and Sadhna Naik.*

### 19.      Relative to peers, Estonia’s tax revenue-to-GDP ratio has been also generally low,

### 19.      Relative to peers, Estonia’s tax revenue-to-GDP ratio has been also generally low

### Tax revenue benchmarking and key comparisons
- By 2022 (last year for which comparable cross-country data is available), total tax collection in Estonia was about 4 pp and 16 pp lower than in the EA20 and the Nordics, respectively.
- Revenue levied through excise duties, VAT and social contributions was broadly comparable or even higher than that of Baltic peers, the Nordics, and the EA20.
- Revenues from personal income taxes, corporate income taxes (CIT), and property taxes were significantly lower than in comparator country groups.
- Estonia ranks almost last in Europe in both corporate income tax and property tax revenue categories.

### Detailed tax-category figures and comparisons (2022)
- Property taxes yield only 0.2 percent of GDP in revenues in Estonia.
  - This is about 4 times less than in the EA20 and the Nordics.
- CIT revenues are 1.7 percent of GDP in Estonia.
  - CIT revenues are four times larger in the Nordics and twice as high in the EA20 (relative comparisons noted in source).
- Personal income tax, VAT, excise, and social contributions:
  - Figures show personal income tax and other consumption/labor taxes often at par or higher than comparator groups, whereas corporate income and property taxes are materially lower.
- Note: Figures on top of selected country data in the source reflect the difference relative to Estonia.

### Conclusions on public finances and policy implications
- Estonia faces a tension between historically low taxation and upward-trending government spending.
  - Emerging spending pressures: defense (recently increasing), long-standing ageing-related spending (health, pensions, old-age benefits), and expectations of broader provision of public services.
- Government revenues have been broadly stable since after the Global Financial Crisis and generally lower than Nordic and European peers.
- Spending profile:
  - Estonia underspends comparators on health, social protection, and environment protection.
  - Estonia outspends the EA20 average on public consumption and investment.
  - Increasing interest expenditures (in the face of larger debt and higher interest rates) will continue to divert resources from other areas.
  - Areas where Estonia outspends comparators—defense, education, and R&D—are unlikely to provide material cost savings given their strategic role.
- Revenue mobilization options:
  - Consumption taxes (VAT and excise duties) and labor taxes (social contributions) are at par or higher than comparator groups.
  - Less distortionary taxation—corporate income and property taxes—yield significantly less government revenue in Estonia than in Nordic and European comparators.
  - Policy implication: along with spending-side consolidation, increasing revenues from underutilized less-distortionary taxes could be an option.

### Banking sector overview and key indicators (selected points from subsequent section)
- Banks represent about 70 percent of the country’s financial assets (Eesti Pank, 2023a).
- Banks’ assets increased from around 100 percent of GDP in 2012 to 136 percent currently.
- Banking sector structure:
  - Comprises nine licensed banks (down from 42 in 1992) and five branches of foreign banks.
  - Four banks are classified as other-systemically important institutions (O-SII).
  - Largest bank by assets: Luminor; second and third: Swedbank and SEB; fourth O-SII: LHV Pank.
- Business model and balance sheet shares:
  - Majority of banks’ assets are loans to households and non-financial corporations (NFCs), constituting 45 percent and 41 percent of total bank lending, respectively.
  - Of the stock of NFC loans, over 42 percent are directed towards the real estate and construction sectors.
  - Loans to government and government bonds are a relatively small portion of banks’ assets.
  - Primary funding source: deposits, which account for approximately 81 percent of liabilities.
  - Banking Sector Balance Sheet Structure, 2023Q2: Loans, 72%; Deposits, incl. CB and credit institutions, 81%; Equity, 11%; Cash at central bank & other, 22%; Debt securities, 4% (assets) and 6% (equity and liabilities) as shown in source figures.
- Macroprudential measures currently in effect (as listed):
  - Countercyclical capital buffer: 1% (1.5% from Dec. 1, 2023).
  - Other systemically important institutions buffer: 2%.
  - Risk weight floor for mortgage loans: 15% (applies to credit institutions using IRB approach; effective as of 2019).
  - Requirements for issuing housing loans (limits may be breached by 15% of the volume of mortgages issued each quarter):
    - Loan-to-value limit (LTV): 85%.
    - Debt service-to-income limit (DSTI): 50%.
      - The calculation of payment must use the interest rate set in the contract plus two percentage points or 6%, whichever is higher. The 2pp add-on is discontinued from Apr. 1, 2024.
    - Maximum maturity: 30 years.
- Financial soundness and performance:
  - System CAR (capital adequacy ratio) is 22.6 percent (among the highest in the EU).
  - Return on equity (RoE) has been consistently around 5–7   percent in recent years—about 2 percentage points higher than in the EU—and improved in 2023.
  - Liquidity coverage ratio (LCR) is over 175 percent (minimum requirement 100 percent).
- Funding and deposits:
  - Non-resident deposits account for about 12 percent of total liabilities in 2023.
  - Deposits sourced through online platforms are 3 percent of total sector deposits.
- Asset quality:
  - Non-performing loans are low at 1.2 percent of total loans (lower than before the pandemic).
  - Sectors with emerging credit-quality deterioration include transport, construction, manufacturing, and some professional services (these sectors account for around 40 percent of banks’ loan books).
- Profitability and interest-rate pass-through:
  - Average interest rate on the stock of corporate loans rose from 2.6 percent in 2021 to 6.2 percent in the third quarter of 2023.
  - Bank profits have increased markedly, driven by rising net interest income, but the surge is largely cyclical.
- Capital adequacy trends and heterogeneity:
  - CET1 ratio fell from over 40 percent in 2014 to 21.5 percent recently.
  - Decline driven in part by growing bank leverage, expansion of loan portfolios, and reforms to corporate taxation that incentivized dividend payouts over profit retention.
  - Supervisory measures and changes in risk-weighted assets—particularly at Swedbank and to a lesser extent SEB—have contributed to declines in aggregate CAR.
  - Small (non-systemically important) banks are, on aggregate, about 4 percentage points less capitalized than the average.

*IMF staff calculations and country data as presented in the source document.*

### 15.      The capital adequacy ratio differs significantly across smaller banks as well. For

### 1estea2024002 - 15.      The capital adequacy ratio differs significantly across smaller banks as well. For

### Capital adequacy variation across banks
- The CET1 ratio of smaller banks ranges between 10.1 and 28.2 percent and the median is significantly lower than for O-SIIs.
- Solvency ratios exhibit significant variation across different banks.
- Overall capital adequacy for the banking system stood at 22 percent currently (pre-counterfactual adjustments).

### Risk Weights on Mortgage and Corporate Lending at IRB Banks: overview
- Two Estonian O-SII banks utilize IRB methodologies for:
  - i) mortgage loans on residential real estate (RRE) to households, and
  - ii) lending to NFCs.
- These two banks account for about three quarters of the total stock of housing loans in Estonia and about one third of the stock of NFC lending.
- IRB-based risk weights have trended downward, reflecting historical loss experiences and a favorable borrowing environment with declining shares of overdue and non-performing loans.
- Eesti Pank introduced a 15 percent floor on the average risk weights for mortgage loans in September 2019.
- No floor currently exists for IRB-based risk weights for NFC loans.
- The standardized methodology, applied by non-IRB banks in Estonia, imposes a risk weight of 35 percent to RRE exposures.
- International comparisons:
  - Sweden: risk weight floors at 25 percent for mortgage exposures on RRE and 35 percent for commercial real estate since 2013.
  - Netherlands: variable floor for RRE that differentiates RW by LTV; RW increases from 12 percent for LTV < 55 percent up to 26.85 percent for LTV = 100 percent.

### Estimated risk weights and comparison with KKR (2024)
- For mortgage loans, IMF staff estimated average IRB-based risk weight: 22 percent; KKR (2024) estimate: 15 percent.
- For NFC loans, IMF staff estimated average IRB-based risk weight: 74 percent; KKR (2024) estimate: 65 percent.
- System-wide averages for risk weights under the standardized approach are comparable across the two analyses.

### Counterfactual exercise 1: 35 percent floor on RRE for IRB banks
- Imposing a 35 percent floor on mortgage loans for the two IRB banks would:
  - Reduce capital ratios by 113bps for Bank 1.
  - Reduce capital ratios by 153bps for Bank 2.
  - Result in a 67bps decrease in the overall capital ratio at the systemic level.

### Counterfactual exercise 2: bringing NFC IRB risk weights to system average
- Calibration assumption: an 85 percent risk weight for NFC loans to align IRB banks with the Estonian banking system average.
  - This 85 percent broadly matches Bank 1’s estimated IRB-based NFC risk weight (83 percent).
  - This 85 percent significantly exceeds Bank 2’s estimated IRB-based NFC risk weight (~55 percent).
- Impact:
  - System-wide impact would be a 144bps reduction in the capital ratio.

### Combined counterfactual impact
- The combined effect of:
  - a 35 percent RW floor on RRE loans for IRB banks, and
  - aligning NFC IRB risk weights with the system average (85 percent),
  would lower the overall capital adequacy ratio of the banking system by approximately 200bps, from 22 percent currently to 20 percent.

### Taxation of extra profits and one-off dividend payout
- Several EU countries have implemented new taxes on banks’ extra profits or raised tax rates; tax design varies significantly across countries.
- In Estonia:
  - No direct tax has been levied on bank profits historically.
  - A large taxable, one-off dividend payout, on top of the ordinary annual dividend distribution, is expected to temporarily support government fiscal revenue but further reduce bank capital ratios by 3 percentage points.
  - For the estimate, it is assumed that the full payout would have accrued to capital, if not paid as dividend.
- Combined effects:
  - The combined effect of the higher calibrated risk weights and the dividend payout would significantly reduce Estonian banks’ capital headroom.
  - A caveat: the dividend payout is not exogenous to capital ratios and may have not occurred had risk weights been higher and capital ratios lower.

### Key findings on banking sector resilience
- Despite a challenging macroeconomic environment, Estonian banks have remained resilient and maintained soundness.
- Capital levels have remained adequate, but there has been a gradual decline in capital adequacy in recent years.
- Estonian banks achieved record profits in 2023, with Non-Performing Loan (NPL) ratios remaining low despite higher interest rates.

### Policy recommendations (banking sector)
- Avoid taxes on windfall profits or initiatives encouraging higher taxable dividend payouts, recognizing:
  - the cyclical nature of the current upswing in bank profits, and
  - the pivotal role bank profits play in bolstering capital buffers during downturns.
- Review bank exposures to ensure that credit risk is properly reflected in risk weights across the banking system.
- Review macro- and micro-prudential requirements for less significant institutions to ensure current regulations promote financial stability uniformly across the banking sector.

### Transition to climate policy section: summary of key points (selected findings)
- Estonia remains one of the most emission-intensive economies in the EU, largely due to reliance on oil shale for about 60 percent of its energy supply.
- Estonia has set targets:
  - climate neutrality by 2050,
  - interim target to cut GHG emissions by 80 percent by 2035 compared with 1990 levels.
- Achievements and gaps:
  - Estonia reduced its GHG emissions by about 42 percent since 2010.
  - At current policies, climate neutrality by 2050 is largely out of reach absent a deep restructuring of the energy mix.
- Electricity grid and dispatchable capacity:
  - Estonia is currently synchronized with Russia and Belarus; interconnector project to connect the Baltic States with continental Europe via Poland by February 2025 is expected to ease constraints.
  - Authorities foresee a lack of dispatchable capacity already in 2027 at current trends in electricity demand.
  - Identified options for additional dispatchable generation include gas and nuclear (Small Modular Reactors).
- Mitigation policy priorities:
  - Gradually raising carbon prices is emphasized as key; Eesti establishments face one of the lowest effective carbon prices in the EU due to low effective ETS coverage.
  - Reducing fossil fuel subsidies—both explicit and implicit—is important. Notable implicit subsidy: undercharging of oil shale producers for pollution and disposal of production waste.
  - Explicit subsidy definition: retail price below supply cost.
  - Implicit subsidy definition: retail price fails to internalize externality cost of emissions and other congestion costs.

*Source: IMF staff report content (Republic of Estonia).*

### 13.      Gradually phasing out oil shale production as planned is essential for achieving the

### 13.      Gradually phasing out oil shale production as planned is essential for achieving the

### Oil shale phase-out: rationale and regional impact
- Oil shale remains Estonia’s largest energy source despite declining use.
- Environmental and resource impacts of oil shale production:
  - One of the most carbon-intensive forms of electricity and heat generation, and of oil production.
  - Generates substantial amounts of solid waste, contributes to air pollution, and requires extensive water usage.
- Policy targets:
  - End oil shale use in electricity production by 2035.
  - End oil shale use in all energy production by 2040.
  - These targets have not been made binding by law, raising the risk of policy reversal.
- Regional dependence and mitigation:
  - Ida-Virumaa region relies on oil shale for 40 percent of its GDP (OECD, 2024).
  - Territorial Just Transition Plan supported with EUR 354 million funding from the EU for investments in renewable energy and reskilling of local workers.

### Greening Estonia’s energy mix
- Renewable energy progress:
  - Share of renewables in energy consumption reached almost 40 percent of total in 2022.
- Required actions to reduce reliance on oil shale:
  - Significant increase in the share of renewable sources.
  - Penalizing the use of oil shale in energy production.
  - Fast tracking investment in wind and solar power generation.
  - Investment in grid integration, storage and dispatchable generation to support the EU ETS in greening Estonia’s energy mix.
  - Targeted subsidies to encourage renewable deployment.

### Taxation, spending, and carbon pricing
- Policy tools to accelerate decarbonization:
  - Carbon taxes, subsidies for renewable energy projects, and incentives for energy-efficient practices.
- Environmental tax collection:
  - Estonia is broadly in line with the EU average for collection of environmental tax revenues.
  - The transport tax component stands significantly below that of EU peers and the EU average, mainly reflecting the lack of a car tax.
- Transport sector emissions drivers:
  - High emissions from transport owe in large part to a relatively energy-inefficient car fleet.
  - A car tax could be designed to guide accelerated improvement in the emission intensity of the car stock over time.

### Sectoral policies: transport, residential, and others
- Transport sector:
  - GHG emissions in the transport sector have increased steadily since the early 1990s, driven by road transport.
  - Increase driven by the number of vehicles—mostly passenger cars—and kilometers driven.
  - Emission efficiency in CO2/km of new passenger cars has been below the EU average, reflecting the low share of zero- and low-emission vehicles.
  - Policy lever: well-calibrated registration and road taxes to promote greater vehicle efficiency and reduce emissions.
- Residential sector:
  - Around 90 percent of Estonia’s residential housing stock is rated below D on the Energy Performance Certificates (EPCs).
  - Thought experiment: a full upgrade of Estonia’s housing stock to the highest energy efficiency standard (EPC=A) would reduce emissions per capita by 42 percent relative to current levels.
  - Country-specific retrofitting cost and energy cost savings: energy cost savings would fully repay investment costs in a shorter time than required on average in the EU, supporting prioritization of related investment.

### Physical risk, adaptation, and resilience
- Observed climate trends and impacts:
  - Estonia has experienced a faster temperature rise than the world average, with steady decline in winter ice and snow coverage.
  - Cumulated losses from extreme weather events have remained contained so far, at less than 1 percent of 2020 GDP over the period 1980-2020, against an EU average of around 3.5 percent of GDP.
- Forward-looking vulnerability and readiness:
  - According to the IMF-Adapted ND-GAIN Country Index, Estonia ranks in the upper quartile of the distribution of 182 countries globally.
  - The INFORM index suggests Estonia’s vulnerability will remain unchanged even under more extreme warming scenarios.
- Adaptation needs:
  - Shifts in precipitation patterns, rising sea levels, and increased frequency of extreme weather events pose risks to infrastructure, the economy, and ecosystems.
  - Ongoing monitoring, evaluation, and proactive planning are essential to address emerging threats and vulnerabilities.

### Conclusions and policy implications
- Mitigating climate change in Estonia requires coordinated policies to:
  - Address barriers such as low effective carbon pricing and high fossil fuel subsidies.
  - Leverage opportunities in renewable energy deployment, taxation, spending, and technology.
- Priorities highlighted:
  - Make phase-out commitments durable to reduce risk of policy reversal.
  - Use EU-supported Territorial Just Transition Plan funding (EUR 354 million) to contain regional disparities and support affected workers and investments.
  - Accelerate renewable investment, grid integration, storage, and dispatchable generation.
  - Consider taxation measures (including a possible car tax) and targeted subsidies to shift incentives toward low-carbon choices.
  - Pursue large-scale building stock renovation to capture substantial emissions reductions (42 percent per capita in the EPC=A thought experiment) with favorable payback relative to EU averages.
- While Estonia exhibits relative resilience to immediate climate risks, proactive adaptation planning is imperative to ensure long-term sustainability.

*Source: IMF staff assessment in “Gradually phasing out oil shale production as planned is essential for achieving the” (chapter content).*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2024/english/1estea2024002.pdf_