## _wp11270

## Source details

**Canonical URL:** [_wp11270](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11270.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11270.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11270.pdf.json)

---

### I. Purpose and key findings
- Purpose: quantify variability of tax elasticities in Lithuania and assess current positions.
- Context:
  - Tax system relies highly on taxing flows (direct and indirect taxes) rather than stocks (wealth and immovable property).
  - Macroeconomic flows (GDP, private consumption, wage bill) are more volatile than the EU average.
- Scope: time-varying elasticities for VAT, PIT, CIT, and excise duties (EX), with emphasis on the recent recession and standardized VAT collections relative to other new EU member countries.
- Key findings:
  - Strong evidence of cyclicality in the elasticity of VAT revenues in Lithuania.
  - Long-run VAT elasticity is close to one.
  - Revenue collections deviated from their long-run equilibrium up to 15 percent over the business cycle.
  - Pre-recession boom period (2006-2008): positive deviation of VAT revenues from long-run equilibrium; turned negative during the bust.
  - As of end-2010, VAT revenues are rebounding but remain about 5 percent short of long-run equilibrium.
  - PIT and CIT elasticities generally exceed unity for most of the sample (progressivity).
  - CIT elasticity exhibits the widest range of variation; EX elasticity follows.
  - All elasticities increased during the recent recovery, with differing intensity.

### II. Literature and sources of elasticity variability
- Methodological approaches summarized:
  - Disaggregated approach: calibrate elasticities of individual tax categories w.r.t. their respective bases; multiply by elasticities of tax bases w.r.t. the output gap.
    - Typical finding: personal and corporate income taxes progressive (elasticity above one); social security contributions regressive (elasticity below one); indirect taxes neutral (elasticity close to one). VAT elasticity often set to unity in some studies.
  - Aggregated approach: calculate elasticities with respect to the output gap for aggregate government revenues.
- Sources of time variation beyond the business cycle:
  - Boom-and-bust cycles of assets, property prices, and commodity prices.
  - Changes in output composition (e.g., private consumption–driven expansion vs. export-driven expansion).
  - Changes in tax compliance (compliance declines during financial crises; taxpayers delay payments, face bankruptcy risk, shift to informal sector).
  - VAT collection efficiency lower when output gap is negative and informal economy expands.
  - Cross-sectional evidence: emerging countries with weaker tax-compliance institutions tend to have higher VAT collection efficiency.

### III. Lithuanian tax system: structure and statistics (2001–2010)
- Aggregate burden and distribution:
  - Overall tax burden: about 30 percent of GDP.
  - Central government: slightly below 50 percent of tax revenues.
  - Local governments: about 12 percent of tax revenues.
  - Remainder to social security and extra budgetary funds.
- Composition (2001-2010):
  - Indirect taxes: 60 percent of total tax revenues and 35 percent of total government revenues.
  - VAT: 38 percent of total tax revenues.
  - Excise duties (EX): 16 percent of total tax revenues.
  - Personal income tax (PIT): 32 percent of total tax revenues.
  - Corporate income tax (CIT): 10 percent of total tax revenues.
  - Other taxes: 4 percent of total.
- Tax structure characteristics:
  - Reliance on taxing income and consumption (flows) makes revenues more vulnerable to fluctuations than taxes on wealth and capital (stocks), which are among the lowest in the EU.
  - CIT and PIT rate changes are easier to implement unilaterally than VAT and excise duties, which face EU harmonization constraints.
- Social security contributions:
  - Counted separately from tax revenues.
  - Importance increased from 2009 when compulsory health insurance contributions became part of social contributions.
  - Contribution rates rose from 9.2 percent in 1999 to 11.6 percent in 2009.
  - 2009 contribution rate of 11.6 percent of GDP corresponds to the average collection in the new EU member countries for the 1999-2009 period.
- Property and land taxes:
  - 0.37 percent of GDP in property and land taxes in 2010.
  - Only commercial property is taxed, with the annual tax set by municipalities in the 0.3-1 percent range.
  - Unimproved land is taxed at 1.5 percent, subject to numerous exemptions and base reductions.
  - No net wealth tax.

### IV. Stylized facts on VAT and C-efficiency
- Time-varying signals:
  - VAT collections declined by 25 percent between 2008-2009, compared to a 14 percent drop in nominal private consumption.
  - VAT collections grew by 12 percent during 2009-2010, compared to a 3 percent decline in private consumption.
  - VAT revenue growth outpaced private consumption also during the pre-recession boom period.
- VAT C-efficiency scores (VAT revenue to personal consumption ratio divided by VAT statutory rates):
  - Median across new EU member countries: 55 percent (country-level example: Estonia median 85 percent).
  - Panel median moved from 57 percent in 1999 to a peak of 72 percent in 2007 and returned to 62 percent in 2010.
  - Dispersion across countries varied over time, with lowest range during the 2007 pre-recession peak.

### V. Data and estimation methods
- Data:
  - Quarterly data on taxes and their bases for the 1999-2010 period.
  - VAT, PIT, CIT, and EX data and tax system changes obtained from the Ministry of Finance.
  - Tax bases (GDP, personal consumption, wage bill, operating surplus) from Statistics of Lithuania.
  - Panel data on VAT revenues and personal consumption for 10 new EU member countries (Bulgaria, Czech Republic, Estonia, Latvia, Lithuania, Hungary, Poland, Romania, Slovenia, Slovakia) from Eurostat.
  - Social contributions excluded due to absence of pre-2004 quarterly data.
- Estimation approaches:
  - Rolling regressions on individual taxes in Lithuania using a fixed moving window of 28 quarters (7 years); year-on-year percentage changes used.
  - Panel data PMG (Pooled Mean Group) estimator on VAT across 10 new EU member countries to estimate pooled long-run elasticity while allowing short-run heterogeneity.

### VI. Rolling-regression results (28-quarter moving window)
- Method: equation (3) estimated with year-on-year differences; elasticity estimates plotted with ± 2 s.d. bounds.
- VAT elasticity:
  - Range: 0.5 and 1.5.
  - Increased at onset of downturn; remained above 1 during recovery.
- PIT elasticity:
  - Range: 0.9 and 1.4.
  - Upward trend 2006-2008, slight decline during recession, rebounded to pre-recession level; above unity for most of sample.
- CIT elasticity:
  - Range: 1 and 4 (widest variation).
  - Declined during recession, stabilized with recovery; wide variation partly due to uneven schedule of CIT payments within the year.
- Excise duty (EX) elasticity:
  - Range: 0 and 1.
  - Rebounding during recovery; may reflect progress in counteracting cross-border smuggling of fuel and cigarettes.

### VII. Panel PMG results on VAT (pooled long-run estimates)
- Model: PMG estimator of Pesaran et al. (1999) applied to VAT revenues adjusted for statutory rate changes (T) and personal consumption (B); pooled long-run elasticity β1 with country-specific speed of adjustment φi and short-run δi.
- Model selection: Hausman tests do not reject poolability of long-run coefficients; PMG favored over MG and FE in both quarterly and annual regressions.
- Long-run private consumption (LR elasticity) coefficients (Table 1 summary):
  - Nominal quarterly: 0.9809***
  - Real quarterly: 0.9976***
  - Nominal annual: 1.0273***
  - Real annual: 1.0746***
- Speed of adjustment coefficients:
  - Nominal quarterly: -0.1067*** (implies about 10 percent of deviation adjusted within a quarter)
  - Real quarterly: -0.0850***
  - Nominal annual: -0.6264*** (implies about 60 percent of deviation adjusted within a year)
  - Real annual: -0.6022***
- Other model statistics:
  - Number of observations: quarterly 470, annual 110.
  - Hausman test p-values (PMG versus MG): 0.5783, 0.3628, 0.2076, 0.3941.
  - Hausman test p-values (PMG versus FE): 0.9598, 0.9746, 0.9504, 0.9645.
- Interpretation:
  - Long-run elasticity close to one consistent with neutral VAT across new EU member countries.
  - Significant negative speed of adjustment indicates deviations from long-run equilibrium are corrected over time.

### VIII. Deviations from long-run equilibrium and cross-country comparison
- Lithuania VAT revenue deviations:
  - Positive deviations during pre-recession boom (2006-2008).
  - Turned negative during recession (2009-2010).
  - Recent quarters: rebound toward long-run equilibrium but remain about five percent below equilibrium.
- Pre-2006 behavior:
  - Revenue gap positive in end-1999 – 2000 while economy still suffering Russian crisis spillovers; private consumption contributed 3.2 ppt on average to GDP growth vs. real GDP growth 2.1 ppt on average.
  - Revenue gap negative in 2003-2004; private consumption contribution 7 ppt vs output growth 8.8 ppt.
- Baltic comparison:
  - Estonia, Latvia, Lithuania all had positive revenue gaps in 2006–08, then negative after the bust.
  - Estonia returned to positive at beginning of 2009; Latvia and Lithuania gaps stayed negative.
  - Latvia experienced largest revenue drop in recession: revenue gap about 20 percent negative at end-2010 (four times larger than Lithuania's 5 percent negative gap).
  - Estonia's better performance attributed to tighter revenue administration and higher C-efficiency scores.

### IX. Conclusions and policy implications
- Empirical evidence indicates variability of tax elasticities in Lithuania.
- Short-run deviations from long-run elasticities can be driven by:
  - Composition effects in tax bases.
  - Cyclical movements in tax compliance (taxpayers put larger weight on potential gains from tax evasion during economic stress).
- Direction of variation differs across taxes; most elasticities were flat during recession and rebounded with the recovery.
- Panel regressions imply recent VAT collections in Lithuania are about 5 percent below long-run equilibrium, indicating room for further VAT revenue improvement in coming months.
- Policy recommendations:
  - Account for variability of tax elasticities when making short-term tax revenue projections.
  - Recognize that deviations of short-term elasticities from long-run levels can be especially pronounced in new EU member countries that rely heavily on taxing flows rather than stocks.
  - Adjust long-run VAT elasticities above unity during periods of rapid economic expansions and contractions.
  - Vary the extent of adjustment across taxes in line with their responsiveness to the business cycle.

*Source: IMF staff analysis (extracted from provided content).*

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

### _wp11270 - References .............................................................................................................

### I. Introduction
- Purpose: quantify variability of tax elasticities in Lithuania and assess current positions.
- Context:
  - Tax system relies highly on taxing flows (direct and indirect taxes) rather than stocks (wealth and immovable property).
  - Macroeconomic flows (GDP, private consumption, wage bill) are more volatile than the EU average.
- Scope: analyze time-varying elasticities for VAT, PIT, CIT, and excise duties (EX), with emphasis on the recent recession and standardized VAT collections relative to other new EU member countries.
- Key findings:
  - Strong evidence of cyclicality in the elasticity of VAT revenues in Lithuania.
  - Long-run VAT elasticity is close to one.
  - Revenue collections deviated from their long-run equilibrium up to 15 percent over the business cycle.
  - Pre-recession boom period (2006-2008): positive deviation of VAT revenues from long-run equilibrium; turned negative during the bust.
  - As of end-2010, VAT revenues are rebounding but remain about 5 percent short of long-run equilibrium.
  - PIT and CIT elasticities generally exceed unity for most of the sample (progressivity).
  - CIT elasticity exhibits the widest range of variation; EX elasticity follows.
  - All elasticities increased during the recent recovery, with differing intensity.

### II. Related literature on estimating tax elasticities
- Methodological approaches:
  - Disaggregated approach (e.g., Girouard and Andre (1995)): calibrate elasticities of individual tax categories with respect to their respective bases; multiply by elasticities of tax bases w.r.t. the output gap to obtain overall tax elasticities.
    - Finding: personal and corporate income taxes are progressive (elasticity above one), social security contributions are regressive (elasticity below one), indirect taxes neutral (elasticity close to one). In the case of the VAT, elasticity is set to unity without conducting estimations.
  - Aggregated approach (e.g., Fedelino et al. (2009), Congressional Budget Office (2009)): calculate elasticities with respect to the output gap for aggregate government revenues.
- Sources of time variation in elasticities beyond the business cycle:
  - Boom-and-bust cycles of assets, property prices, and commodity prices (Aydin (2010) example: South Africa).
  - Changes in output composition (Bornhorst et al., 2011): e.g., private consumption–driven expansion has larger tax impact than export-driven expansion.
  - Changes in tax compliance (Brondolo (2009)): compliance declines during financial crisis; taxpayers delay payments, face bankruptcy risk, shift to informal sector.
  - VAT collection efficiency lower in “bad” times when output gap is negative and informal economy expands (Sancak et al. (2010)).
  - Cross-sectional evidence: emerging countries with weaker tax-compliance institutions tend to have higher VAT collection efficiency (Agha and Haughton, 1996; De Melo, 2009; Aizenman and Jinjarak, 2008).

### III. Brief overview of the Lithuanian tax system
- Aggregate burden and distribution:
  - Overall tax burden: about 30 percent of GDP (lower than EU average).
  - Proportion of tax revenues: central government slightly below 50 percent; local governments about 12 percent; remainder to social security and extra budgetary funds.
- Composition (2001-2010):
  - Indirect taxes comprised 60 percent of total tax revenues and 35 percent of total government revenues.
  - VAT comprises 38 percent of total tax revenues.
  - Excise duties (EX) comprise 16 percent of total tax revenues.
  - Personal income tax (PIT) comprises 32 percent of total tax revenues.
  - Corporate income tax (CIT) comprises 10 percent of total tax revenues.
  - Other taxes constitute the remaining 4 percent of the total.
- Figure references (as presented in source):
  - Figure 1: Trends in Tax Revenue Collections over the Last Decade (shows CIT, PIT, VAT, Excise duties as share in Total Tax Revenues (Percent) for 2001–2010; and Direct taxes, Indirect taxes, Social contributions as Share in Total Government Revenues (Percent) for 2001–2010).
- Tax structure characteristics:
  - Relies on taxing income and consumption (flows); taxes on wealth and capital (stocks) are among the lowest in the EU, increasing vulnerability to economic fluctuations.
  - Main taxes: PIT, CIT, VAT, EX; characteristics broadly correspond to those elsewhere in the EU.
  - CIT and PIT rate changes are easier to implement unilaterally than VAT and excise duties, which face tougher EU harmonization constraints.
- Social security contributions:
  - Counted separately from tax revenues.
  - Importance increased from 2009 when compulsory health insurance contributions became part of social contributions.
  - Contribution rates rose from 9.2 percent in 1999 to 11.6 percent in 2009.
  - 2009 contribution rate of 11.6 percent of GDP corresponds to the average collection in the new EU member countries for the 1999-2009 period.
- Property and land taxes: revenues are relatively modest (text cuts off at this point in source).

* _wp11270 - References ............................................................................................................. *

### 0.37  percent  of  GDP  in  property  and  land  taxes  in  2010.  At  present,  only  commercial

### _wp11270 - 0.37  percent  of  GDP  in  property  and  land  taxes  in  2010.  At  present,  only  commercial

### IV. STYLIZED FACTS
- Lithuania recorded "0.37  percent  of  GDP  in  property  and  land  taxes  in  2010."
- At present, only commercial property is taxed in Lithuania, with the annual tax being set by the municipalities in the "0.3-1 percent" range.
- Unimproved land is taxed at "1.5  percent", but numerous exemptions and base reductions apply narrowing the taxable base substantially.
- There is no net wealth tax.
- Prima facie evidence suggests time-varying tax elasticities in Lithuania:
  - During the recession VAT collections declined by "25 percent" between 2008-2009, compared to a "14 percent" drop in nominal private consumption.
  - During 2009-2010 VAT collections grew by "12 percent", compared to a "3 percent" decline in private consumption.
  - VAT revenue growth outpaced private consumption also during the pre-recession boom period.
- VAT C-efficiency scores (VAT revenue to personal consumption ratio divided by VAT statutory rates) show:
  - Wide variation across new EU member countries (median score is "55 percent"; Estonia median score is "85 percent").
  - Median of VAT C-efficiency across the panel moved from "57 percent" in 1999 to a peak of "72 percent" in 2007 and returned to "62 percent" in 2010.
  - Dispersion across countries varied over time, with lowest range during 2007 pre-recession peak.

### V. DATA AND ESTIMATION RESULTS
- Data:
  - Quarterly data on taxes and their bases for the "1999-2010" period.
  - VAT, PIT, CIT, and EX data and tax system changes obtained from the Ministry of Finance.
  - Tax bases (GDP, personal consumption, wage bill, operating surplus) from Statistics of Lithuania.
  - Panel data on VAT revenues and personal consumption for 10 new EU member countries (Bulgaria, Czech Republic, Estonia, Latvia, Lithuania, Hungary, Poland, Romania, Slovenia, Slovakia) from Eurostat.
- Two methodologies used to assess variability of tax elasticities:
  - Rolling regression methods on individual taxes in Lithuania (fixed moving window of "28 quarters (7 years)"; year-on-year percentage changes used).
    - Advantage: compares elasticities across subsamples.
    - Drawback: limited precision due to short time series.
  - Panel data methods (pooled mean group estimator) on VAT collections across 10 new EU member countries.
    - Advantage: expands observations, compares performance across countries, captures VAT's share of revenues.
    - Limitation: cannot be applied to PIT, CIT, and EX due to cross-country tax system differences.
- Analysis excluded social contributions due to absence of pre-2004 quarterly data.

### Rolling regressions (results from equation (3) using a moving window of 28 quarters)
- Method:
  - Equation (3) estimated with year-on-year differences to avoid reliance on HP filter results.
  - Elasticity estimates plotted with ± 2 s.d. bounds.
- Findings (time-varying elasticities):
  - VAT elasticity ranged between "0.5 and 1.5".
    - Increased at onset of downturn; remained above 1 during recovery fueling recent tax collection recovery.
  - PIT elasticity ranged between "0.9 and 1.4".
    - Upward trend between 2006-2008, slight decline during recession, rebounding to pre-recession level; stayed above unity for most of sample consistent with progressivity.
  - CIT elasticity ranged between "1 and 4" (widest variation).
    - Declined during recession, stabilized with recovery; wider variation explained by uneven schedule of CIT payments within the year.
  - Excise duty elasticity ranged between "0 and 1".
    - Rebounding during recovery; may reflect progress in counteracting cross-border smuggling of fuel and cigarettes.

### Panel data regressions (PMG estimator on VAT)
- Model specification:
  - PMG (Pooled Mean Group) estimator of Pesaran et al. (1999) applied to VAT revenues adjusted for statutory rate changes (T) and personal consumption (B).
  - Equation (4) estimated: pooled long-run elasticity β1, country-specific speed of adjustment φi, short-run δi, fixed effect αi.
  - PMG imposes homogeneous long-run relationship, allows short-run heterogeneity.
- Model selection:
  - Hausman tests do not reject poolability of long-run coefficients (PMG favored over MG and FE in both quarterly and annual regressions).
  - Suggests long-run VAT elasticity of unity holds for all new EU member countries.
- PMG estimation results (Table 1 summary):
  - Long-run private consumption (LR elasticity) coefficients:
    - Nominal quarterly: "0.9809***"
    - Real quarterly: "0.9976***"
    - Nominal annual: "1.0273***"
    - Real annual: "1.0746***"
  - Speed of adjustment coefficients:
    - Nominal quarterly: "-0.1067***"
      - Implies about "10 percent" of deviation adjusted within a quarter.
    - Real quarterly: "-0.0850***"
    - Nominal annual: "-0.6264***"
      - Implies about "60 percent" of deviation adjusted within a year.
    - Real annual: "-0.6022***"
  - Changes in private consumption (short-run coefficients) reported with statistical significance (see table).
  - Number of observations: quarterly "470", annual "110".
  - Hausman test p-values (PMG versus MG): "0.5783", "0.3628", "0.2076", "0.3941".
  - Hausman test p-values (PMG versus FE): "0.9598", "0.9746", "0.9504", "0.9645".
- Interpretation:
  - Long-run elasticity close to one consistent with neutral VAT.
  - Significant negative speed of adjustment indicates deviations from long-run equilibrium are corrected over time.
  - Deviations during recession provide insight into cyclicality of VAT revenues in Lithuania.

### Deviations of VAT revenues from long-run equilibrium (Figure 5 findings)
- VAT revenue deviations closely tied to economic cycle:
  - Positive deviations during pre-recession boom ("2006-2008").
  - Turned negative during recession ("2009-2010").
  - Recent quarters show rebound toward long-run equilibrium but remain about "five percent below the equilibrium".
- Pre-2006 behavior:
  - Revenue gap positive in end-"1999 – 2000" while economy still suffering Russian crisis spillovers; private consumption contributed "3.2 ppt" on average to GDP growth vs. real GDP growth "2.1 ppt" on average.
  - Revenue gap negative in "2003-2004"; private consumption contribution "7 ppt" vs output growth "8.8 ppt".
- Comparison with other Baltic countries:
  - All three Baltic countries (Estonia, Latvia, Lithuania) had positive revenue gaps in "2006–08", then negative after the bust.
  - Estonia returned to positive at beginning of "2009"; Latvia and Lithuania gaps stayed negative.
  - Latvia experienced largest revenue drop in recession: revenue gap about "20 percent" negative at end-"2010" (four times larger than Lithuania's "5 percent" negative gap).
  - Estonia's better performance attributed to tighter revenue administration and higher C-efficiency scores.

### VI. CONCLUSION (policy-relevant findings and implications)
- Empirical evidence indicates variability of tax elasticities in Lithuania.
- Short-run deviations from long-run elasticities can be driven by:
  - Composition effects in tax bases.
  - Cyclical movements in tax compliance (taxpayers put larger weight on potential gains from tax evasion during economic stress).
- Direction of variation differs across taxes; most elasticities were flat during recession and rebounded with the recovery.
- Panel regressions imply recent VAT collections in Lithuania are about "5 percent" below long-run equilibrium, indicating room for further VAT revenue improvement in coming months.
- Policy implications:
  - Variability of tax elasticities should be accounted for when making short-term tax revenue projections.
  - Deviations of short-term elasticities from long-run levels can be especially pronounced in new EU member countries that rely heavily on taxing flows (rather than stocks), which are more volatile than EU average.
  - Long-run VAT elasticities should be adjusted above unity during periods of rapid economic expansions and contractions.
  - The extent of adjustment should vary across taxes in line with their responsiveness to the business cycle.

*Source: IMF staff analysis (extracted from provided content)._*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11270.pdf_