## Recent Trends of Informality in Greece – Evidence from Subnational Data

## Source details

**Canonical URL:** [Recent Trends of Informality in Greece – Evidence from Subnational Data](https://www.imf.org/-/media/files/publications/selected-issues-papers/2024/english/sipea2024009.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/selected-issues-papers/2024/english/sipea2024009.pdf.md)
- [Structured JSON version](/-/media/files/publications/selected-issues-papers/2024/english/sipea2024009.pdf.json)

---

### Literature and Approaches
- Informality definitions and estimation:
  - Informality is variably defined (e.g., “shadow economy”, “grey economy”) and estimated using micro survey data or model-based approaches.
- Key prior findings referenced:
  - Elgin and Schneider (2016): studied levels and driving forces of informality in 38 OECD countries using both DGE and MIMIC approaches.
  - Elgin and others (2019): share of the informal economy for advanced economies declined to an average of 17 percent in 2016.
  - Hu and Yao (2022): satellite nightlight data can improve estimates of true GDP per capita.
- Greece-specific evidence from earlier studies:
  - Artavanis and others (2016): estimated about 43-45 percent of self-employment income was not reported.
  - Dellas and others (2017): using a DSGE model, informality increased from 25 percent of GDP to between 35 and 40 percent of GDP during 2010–15.
  - Kelmanson and others (2019): informality in Greece declined from 32 percent of GDP in 2009 to 30 percent of GDP in 2016.
  - Schneider and Asllani (2022): informality in Greece declined from about 24 percent in 2013 to 19 percent in 2019 but edged up in 2020 and 2022.
- Data and method advances:
  - Recent progress in big data and subnational data (including satellite nightlight and Google searches) offers better support for subnational analysis.

### Methodology and Data
- Estimation approach:
  - MIMIC method adopted, following Medina and Schneider (2018), with nightlight intensity data mitigating endogeneity bias.
  - MIMIC links multiple observable indicators and multiple observable causes through a latent informality index estimated by maximum-likelihood and calibrated to survey data.
- Frequency and coverage:
  - Subnational quarterly data for 2012–2021 used, including satellite nightlight and Google searches, complementing conventional indicators.
- Observable indicators:
  - Satellite nighttime lights: VIIRS Day/Night radiance; sum of nighttime light intensity within each region used as proxy for economic activity.
  - Google search: search volume of the name of a subnational region used as proxy for tourism-related economic activity.
  - Conventional variables: labor participation rate (population aged 20-64) and GDP at NUTS 2 subnational level; both transformed into first differences.
- Causes included (subnational and national):
  - Subnational: agriculture employment share, self-employment share, unemployment rate, lagged income per capita, tourist arrivals as share of population, education attainment (share of population aged 25-64 with less than primary, primary, and lower secondary education).
  - National: VAT gap, tax as percent of GDP, trade openness, government consumption as percent of GDP, World Bank governance indicators.
- Calibration:
  - World Bank Enterprise Survey for Greece (2018) at NUTS 1 level used to calibrate the index; each NUTS 2 region assigned the same level of informality as its NUTS 1 region.
  - Key survey variables:
    - e: “Does this establishment compete against unregistered or informal firms?” (0/1)
    - l: “What is the number of permanent, full-time employees at the end of last fiscal year?”
  - Informal-size calculation described:
    - number of employees in firms competing with informal firms is ∑ e_k l_k
    - size of the informal economy computed as z_k =  ∑e_k l_k / ∑e_k l_k + ∑ l_k

### Main Findings: Trends and Magnitudes
- Regional and temporal trends (2010s–2021):
  - Informality declined significantly in all regions over the past decade, with brief increases around 2013–15 and 2020.
  - During the sovereign debt crisis, most regions experienced increased informality by an average of 0.6 percentage point, peaking at about 30 percent of GDP at the national level.
  - Regional variation in declines ranges from 5 percent of GDP (Western Macedonia) to 18 percent of GDP (Crete).
  - Drops above 15 percent of GDP observed in Attica and island regions (Crete, Ionia Islands).
  - Western Macedonia saw higher increases around the sovereign crisis and a more moderate decline thereafter.
  - Western Greece saw a smaller decline at about 10 percent of GDP.
- National aggregate and tax implications:
  - Aggregate informality (weighted by regional gross value added) declined from a peak of about 30 percent of GDP in 2013 to a low of about 16 percent of GDP in 2021.
  - This decline is associated with a gain in tax revenue by about 4 percentage points of GDP (national-level buoyancy approach).
  - EC estimates: Greece’s Value Added Tax gap decreased by over 16 percentage point of its tax base between 2013 and 2021, or about 2 percentage points of GDP.
  - Lower informality is correlated with higher subsequent per capita GDP growth.

### Digitalization Correlations (Estimated Effects)
- Internet and informality associations (average effects reported):
  - A 10-percentage point increase in internet access is on average associated with a reduction of informality by about 3 percentage point.
  - A 10-percentage point reduction in the share of residents who have never used internet is associated with a reduction in informality by about 4 percentage points.
  - A 10-percentage point increase in the share of residents who used online purchase is associated with a reduction in informality by about 4 percentage points.
  - A 10-percentage point increase in the share of residents who used digital public services is associated with a reduction in informality by about 3 percentage points.
- Interpretation:
  - Results are consistent with prior findings that digital payments reduced informality and tax evasion.

### Policy Discussions and Recommendations
- Factors contributing to the decline in informality:
  - Tightened regulation and enforcement against financial crimes and tax evasion.
  - Increased labor market flexibilities (e.g., working hours) that reduced incentives to remain informal.
  - Progress in digitalization: improved infrastructure, digital public services, and modernized employment information systems that eased registration and increased monitoring.
  - Policy continuity, political stability, and reduced corruption may have improved tax morale.
  - Rationalization of tax policies, such as reducing business income tax rates and marginal personal income tax, likely lowered incentives for tax evasion.
- Potential further actions recommended:
  - Strengthen digital infrastructure and digital skill training to address remaining gaps and further reduce informality.
  - Integrate information systems to enhance monitoring of tax compliance, ease business and employment registration, and better integrate the self-employed into the formal economy.
  - Leverage National Resilience and Recovery Plan and Next Generation EU funding to further digitalization and related skill training.
- Cross-country and structural recommendations highlighted:
  - Improving regulatory quality and transparency to ease the burden for business entry and support more productive competition.
  - Increasing efficiency in spending on education and training to increase industry-relevant skills, job matching, and incentives for formal employment.
  - More targeted employment support for youth and women with disproportionately elevated unemployment rates.
  - Reforms to better link firm and job formalization with access to finance to enhance incentives to formalize.
  - Concerted efforts to integrate high self-employment activities into the formal economy given Greece’s relatively high level of self-employment.

### MIMIC Model: Structure, Estimation, and Key Results
- Model structure (summary):
  - Structural equation: y_k^* = α′ x_k + η_k, where η_k captures components not explained by causes x_k.
  - Measurement equation: y_k = β y_k^* + ε_k.
  - Reduced-form: y_k = β α′ x_k + (β η_k + ε_k).
  - Error covariance: Ω = E[(β η_k + ε_k)(β η_k + ε_k)′] = σ^2 β β′ + Θ^2.
  - Normalization: assume first indicator has same unit as y^*, y_1k = y_k^* + ε_1k.
  - Estimation: maximum-likelihood estimation.
- Data transformation and calibration steps:
  - Country fixed effects added by demeaning variables: y_iitjk − ȳ_ij where ȳ_ij is the average of the i-th indicator for country j over all time periods; same applied to cause variables.
  - From index to size calibration: z_k = λ s_k + δ with λ = (z_k2 − z_k1) / ∑_{k1}^{k2} ŷ_k^* and δ = z_k1 − λ s_k1; with more points λ and δ can be estimated by OLS.
- Selected MIMIC estimation results for Greece (coefficients and statistics preserved exactly):
  - Structural equation coefficients (variable — coefficient std.err. z p-value):
    - agriculture employment share (%) 0.47 0.12 4.02 0.00
    - self employment share (%) 0.16 0.11 1.51 0.13
    - unemployment rate (%) 0.27 0.08 3.42 0.00
    - tourist arrivals as mutiple of population -0.01 0.00 -6.24 0.00
    - (lagged) GDP per capita 0.32 0.03 10.63 0.00
    - population aged 25-64 with secondary education or lower (%) -0.67 0.09 -7.90 0.00
    - VAT gap 0.59 0.05 11.74 0.00
    - tax (% of GDP) -2.16 0.32 -6.74 0.00
    - trade openness -0.46 0.05 -9.35 0.00
    - government consumption (% of GDP) 1.14 0.25 4.52 0.00
    - rule of law -0.08 0.02 -3.76 0.00
    - control of corruption 0.02 0.02 1.31 0.19
    - government effectivenss -0.03 0.02 -1.35 0.18
    - political stability 0.16 0.02 7.74 0.00
  - Measurement equation coefficients (indicator — Informal coefficient std.err. z p-value):
    - (-) GDP growth Informal 1 (constrained)
    - labor participation rate growth Informal -0.14 0.01 -9.79 0.00
    - Google Search Volume growth Informal -0.13 0.53 -0.24 0.81
    - Nighttime light growth Informal -0.07 0.10 -0.73 0.47

### Caveats and Directions for Future Research
- Limitations noted:
  - If relations between variables (including satellite nightlight intensity or Google search) and formal/informal activity diverge over time, such effects are not fully controlled.
  - Lack of more relevant subnational data (e.g., non-cash transactions) constrains tighter linkage to regional economic activities.
  - Social structure impacts across Greek regions may have strong effects on informality and require research beyond this paper’s scope.
- Suggested future research directions:
  - Better calibration using more available Greece firm surveys.
  - Incorporation of additional subnational indicators (non-cash transactions, finer firm-level data).
  - Deeper analysis of drivers of subnational variation in informality.

*IMF Selected Issues Papers — Recent Trends of Informality in Greece – Evidence from Subnational Data*

### 3.      There is a large literature on estimating informality including for EU countries, albeit

### Recent Trends of Informality in Greece – Evidence from Subnational Data

### Literature and Approaches
- Informality is variably defined in the literature (e.g., “shadow economy”, “grey economy”) and estimated using micro survey data or model-based approaches.
- Key findings from prior studies:
  - Elgin and Schneider (2016): studied levels and driving forces of informality in 38 OECD countries using both DGE and MIMIC approaches.
  - Elgin and others (2019): share of the informal economy for advanced economies declined to an average of 17 percent in 2016.
  - Hu and Yao (2022): satellite nightlight data can improve estimates of true GDP per capita.
- Greece-specific evidence:
  - Artavanis and others (2016): estimated about 43-45 percent of self-employment income was not reported.
  - Dellas and others (2017): using a DSGE model, informality increased from 25 percent of GDP to between 35 and 40 percent of GDP during 2010–15.
  - Kelmanson and others (2019): informality in Greece declined from 32 percent of GDP in 2009 to 30 percent of GDP in 2016.
  - Schneider and Asllani (2022): informality in Greece declined from about 24 percent in 2013 to 19 percent in 2019 but edged up in 2020 and 2022.
- Recent progress in big data and subnational data offers better support for subnational analysis.

### Methodology and Data
- Estimation approach:
  - MIMIC method adopted, with improvements in data and estimation methods, following Medina and Schneider (2018) who found MIMIC advantageous and that nightlight intensity data mitigates endogeneity bias.
  - MIMIC links multiple observable indicators and multiple observable causes through a latent informality index (Annex I). The latent index is estimated by regression with restrictions via maximum-likelihood estimation and then calibrated to survey data to relate to formal GDP statistics.
- Added data sources and frequency:
  - Subnational quarterly data for 2012–2021, including satellite nightlight and google searches, to complement conventional indicators.
- Observable indicators used:
  - Satellite nighttime lights: VIIRS Day/Night radiance; sum of nighttime light intensity within each region used as proxy for economic activity.
  - Google search: search volume of the name of a subnational region used as proxy for tourism-related economic activity.
  - Conventional variables: labor participation rate (population aged 20-64) and GDP at NUTS 2 subnational level; both transformed into first differences.
- Causes included:
  - Subnational: agriculture employment share, self-employment share, unemployment rate, lagged income per capita, tourist arrivals as share of population, education attainment (share of population aged 25-64 with less than primary, primary, and lower secondary education).
  - National: VAT gap, tax as percent of GDP, trade openness, government consumption as percent of GDP, World Bank governance indicators.
- Calibration:
  - World Bank Enterprise Survey for Greece (2018) at NUTS 1 level used to calibrate the index. Each NUTS 2 region assigned the same level of informality as its NUTS 1 region.
  - Key survey questions used:
    - “Does this establishment compete against unregistered or informal firms?” (e: 0/1)
    - “What is the number of permanent, full-time employees at the end of last fiscal year?” (l)
  - Calculation described: number of employees in firms competing with informal firms is ∑ e_k l_k; size of the informal economy computed as
    - z_k =  ∑e_k l_k / ∑e_k l_k + ∑ l_k

### Main Findings
- Regional trends (2010s–2021):
  - Informality declined significantly in all regions over the past decade, with brief increases around 2013–15 and 2020.
  - During the sovereign debt crisis, most regions experienced increased informality by an average of 0.6 percentage point, peaking at about 30 percent of GDP at the national level.
  - Variation in regional declines ranges from 5 percent of GDP (Western Macedonia) to 18 percent of GDP (Crete).
  - Drops above 15 percent of GDP observed in Attica and island regions (Crete, Ionia Islands). Western Macedonia saw higher increases around the sovereign crisis and a more moderate decline thereafter. Western Greece saw a smaller decline at about 10 percent of GDP.
- National aggregate and tax implications:
  - Aggregate informality (weighted by regional gross value added) declined from a peak of about 30 percent of GDP in 2013 to a low of about 16 percent of GDP in 2021.
  - This decline is associated with a gain in tax revenue by about 4 percentage points of GDP (national-level buoyancy approach).
  - EC estimates: Greece’s Value Added Tax gap decreased by over 16 percentage point of its tax base between 2013 and 2021, or about 2 percentage points of GDP.
  - Lower informality is correlated with higher subsequent per capita GDP growth.
- Digitalization correlations:
  - A 10-percentage point increase in internet access is on average associated with a reduction of informality by about 3 percentage point.
  - A 10-percentage point reduction in the share of residents who have never used internet is associated with a reduction in informality by about 4 percentage points.
  - A 10-percentage point increase in the share of residents who used online purchase is associated with a reduction in informality by about 4 percentage points.
  - A 10-percentage point increase in the share of residents who used digital public services is associated with a reduction in informality by about 3 percentage points.
  - Results are consistent with prior findings that digital payments reduced informality and tax evasion.

### Policy Discussions and Recommendations
- Factors that contributed to the decline in informality:
  - Tightened regulation and enforcement against financial crimes and tax evasion.
  - Increased labor market flexibilities (e.g., working hours) that reduced incentives to remain informal.
  - Progress in digitalization: improved infrastructure, digital public services, and modernized employment information systems that eased registration and increased monitoring.
  - Policy continuity, political stability, and reduced corruption may have improved tax morale.
  - Rationalization of tax policies, such as reducing business income tax rates and marginal personal income tax, likely lowered incentives for tax evasion.
- Potential for further action:
  - Strengthen digital infrastructure and digital skill training to address remaining gaps and further reduce informality.
  - Integrate information systems to enhance monitoring of tax compliance, ease business and employment registration, and better integrate the self-employed into the formal economy.
  - Leverage National Resilience and Recovery Plan and Next Generation EU funding to further digitalization and related skill training.

### Caveats and Directions for Future Research
- Limitations of the current study:
  - If relations between variables (including satellite nightlight intensity or google search) and formal/informal activity diverge over time, such effects are not fully controlled.
  - Lack of more relevant subnational data (e.g., non-cash transactions) constrains tighter linkage to regional economic activities.
  - Social structure impacts across Greek regions may have strong effects on informality and require research beyond this paper’s scope.
- Future research suggestions:
  - Better calibration using more available Greece firm surveys.
  - Incorporation of additional subnational indicators (non-cash transactions, finer firm-level data).
  - Deeper analysis of drivers of subnational variation in informality.

*IMF Selected Issues Papers — Recent Trends of Informality in Greece – Evidence from Subnational Data*

### 16.      Moreover, cross-country studies underscore the importance of other structural

### 16.      Moreover, cross-country studies underscore the importance of other structural

### Key policy findings and recommendations
- Improving regulatory quality and transparency would ease the burden for business entry and support more productive competition. (OECD (2023), Medina and Schneider (2018), Ohnsorge and Yu (2022), Schneider and Asllani (2022))
- Increasing the efficiency in spending on education and training would increase industry-relevant skills, job matching, and the incentive for formal employment. (Kelmanson and others (2019))
- More targeted employment support is recommended, especially for youth and women who still face a disproportionately elevated unemployment rates in Greece.
- Reforms to better link firm and job formalization with access to finance can enhance the incentive to formalize. (Ohnsorge and Yu (2022))
- Given the relatively high level of self-employment in Greece, concerted efforts are warranted to ensure that corresponding activities are appropriately integrated in the formal economy. (Schneider and Asllani (2022))

### MIMIC model: structure and estimation
- Modeling approach: Multiple Indicator Multiple Cause (MIMIC) model links multiple observable indicators and multiple observable causes to a latent variable (index of the informal economy); index can be transformed and calibrated to compute size of informal economy.
- Structural equation:
  - y_k^* = α′ x_k + η_k
  - η_k is a structural disturbance capturing components not explained by causes x_k.
- Measurement equation:
  - y_k = β y_k^* + ε_k
- Assumptions on disturbances:
  - E(η_k ε_k′ ) = 0′
  - E(η_k^2) = σ^2
  - E(ε_k ε_k′ ) = Θ^2
- Reduced-form representation:
  - y_k = β α′ x_k + (β η_k + ε_k)
- Model restrictions:
  - Π = β α′ has rank one.
  - Error covariance Ω = E[(β η_k + ε_k)(β η_k + ε_k)′] = σ^2 β β′ + Θ^2 (sum of a rank-one matrix and a diagonal matrix).
- Normalization:
  - To pin down α and y^*, normalize by assuming first indicator has same unit as y^*: y_1k = y_k^* + ε_1k.
- Estimation method:
  - The MIMIC model can be estimated by the maximum-likelihood estimation.

### Data transformation and calibration steps
- To control for institutional and cultural differences in subnational data, country fixed effects are added by demeaning all variables at the country level:
  - Each indicator y_iitjk is transformed to y_iitjk − ȳ_ij, where ȳ_ij is the average of the i-th indicator for country j over all time periods. The same transformation is applied to cause variables.
- From index to size:
  - Once α is estimated as α̂, the index is ŷ_k^* = α̂′ x_k and has same unit as y_1k due to normalization.
  - Rescaling and re-calibration are needed because the unit of y_1k (percentage point or percent change) does not always imply direct interpretation as size.
  - Let z_k be size of informal economy. If ŷ_k^* has unit of percentage point and s_k = ∑_{k1}^{k} ŷ_k^*, and if z_k1 and z_k2 are known at two points, calibrate:
    - z_k = λ s_k + δ, where
    - λ = (z_k2 − z_k1) / (s_k2 − s_k1) = (z_k2 − z_k1) / ∑_{k1}^{k2} ŷ_k^*,
    - δ = z_k1 − λ s_k1.
  - With more data points, λ and δ can be estimated by Ordinary Least Squares regression.

### MIMIC estimation results for Greece (selected coefficients and statistics)
- Structural equation coefficients (variable — coefficient std.err. z p-value):
  - agriculture employment share (%) 0.47 0.12 4.02 0.00
  - self employment share (%) 0.16 0.11 1.51 0.13
  - unemployment rate (%) 0.27 0.08 3.42 0.00
  - tourist arrivals as mutiple of population -0.01 0.00 -6.24 0.00
  - (lagged) GDP per capita 0.32 0.03 10.63 0.00
  - population aged 25-64 with secondary education or lower (%) -0.67 0.09 -7.90 0.00
  - VAT gap 0.59 0.05 11.74 0.00
  - tax (% of GDP) -2.16 0.32 -6.74 0.00
  - trade openness -0.46 0.05 -9.35 0.00
  - government consumption (% of GDP) 1.14 0.25 4.52 0.00
  - rule of law -0.08 0.02 -3.76 0.00
  - control of corruption 0.02 0.02 1.31 0.19
  - government effectivenss -0.03 0.02 -1.35 0.18
  - political stability 0.16 0.02 7.74 0.00
- Measurement equation coefficients (indicator — Informal coefficient std.err. z p-value):
  - (-) GDP growth Informal 1 (constrained)
  - labor participation rate growth Informal -0.14 0.01 -9.79 0.00
  - Google Search Volume growth Informal -0.13 0.53 -0.24 0.81
  - Nighttime light growth Informal -0.07 0.10 -0.73 0.47

*IMF Selected Issues Papers: Recent Trends of Informality in Greece – Evidence from Subnational Data*

---


_Source: https://www.imf.org/-/media/files/publications/selected-issues-papers/2024/english/sipea2024009.pdf_