## _wp14110

## Source details

**Canonical URL:** [_wp14110](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14110.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14110.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14110.pdf.json)

---

### Context and key questions
- Many OECD countries face significant fiscal challenges after the crisis, with fiscal deficits having "skyrocketed" and ageing populations putting further pressure on public finances.
- Tax buoyancy measures how tax revenues vary with changes in GDP: a buoyancy of one implies an extra percent of GDP raises tax revenue by 1 percent (keeping the tax-to-GDP ratio unchanged); buoyancy > 1 raises revenue by more than GDP and can reduce the deficit ratio.
- Short-run buoyancy relates to fiscal stabilization (automatic stabilizers); long-run buoyancy matters for fiscal sustainability.
- Key questions:
  - Does observed buoyancy > 1 during some years hold over longer periods?
  - Does buoyancy > 1 hold for all countries and for all taxes?

### Methodology: specification and estimation
- Model: panel autoregressive distributed lag, transformed into an Error Correction Model (ECM) with p=1 and q=1 selected by the Bayesian information criterion.
- ECM parameters of interest:
  - θi,t=1 (short-term parameter) measures the instantaneous effect of a change in GDP on tax revenue (short-term buoyancy).
  - βi (long-run buoyancy) measures the long-run relationship between taxes and GDP.
  - λi (speed of adjustment) measures how fast buoyancy converges to its long-run equilibrium value.
- Estimation approaches:
  - Aggregate revenue: estimated per country using the mean group (MG) estimator; simple average of country estimates yields an OECD average.
  - Revenue by tax category and by period: estimated using panel regressions with the pooled mean group (PMG) estimator, which constrains long-run coefficients to be equal across groups while allowing intercepts and short-run coefficients to differ.
- The ECM framework permits simultaneous estimation of short-run effects, long-term relationship, and speed of adjustment, enabling decomposition of buoyancy into short-run and long-run components.

### Theoretical expectations
- Benchmark long-run buoyancy expected to be one.
- Long-run buoyancy likely larger for more progressive taxes (e.g., PIT) and smaller for regressive taxes (e.g., SSC, GST).
- VAT/GST buoyancy may exceed one if luxury goods (income elasticity > 1) are largely taxed at standard rates while necessities face reduced rates.
- Structural changes (e.g., decline in labor-income share) can reduce SSC buoyancy and increase CIT buoyancy.
- Excise taxes not indexed to GDP growth can yield buoyancy < 1.
- Short-run expectations by tax:
  - CIT expected to have relatively high short-run buoyancy (good automatic stabilizer).
  - Property taxes may have particularly small short-run buoyancy if local governments stabilize them counter-cyclically.
  - PIT and SSC may have short-run buoyancy below one in countries with rigid wages and tight employment protection.
  - GST short-run buoyancy can be below one due to consumption smoothing.
  - Recessions can reduce compliance and lower short-term buoyancy.

### Data
- Annual tax revenue data for 34 OECD countries, 1965–2012, from OECD revenue statistics.
- For countries joining OECD after 1965, data begin at accession; "The first year of data for the whole sample is 1995."
- Number of years per country varies from 16 for Estonia to 47 for the oldest members.
- GDP data from IMF sources.
- In 2012, the average tax-to-GDP ratio in the OECD is 34 percent.
- Revenue categories used: PIT, CIT, SSC, GST, Excises, and Property taxes.
- Observed composition dynamics (OECD average, balanced panel since 1965): steady increase in GST and SSC shares; PIT rose until late 1980s then started to fall. PIT and SSC are the two main sources of revenues.

### Results — Buoyancy of Total Tax Revenue (aggregate, country-by-country)
- Country-level estimation of the ECM for total tax revenue across 34 OECD countries yields:
  - Average long-run buoyancy: 1.03
  - Median long-run buoyancy: 1.05
- Significance across countries:
  - Long-run buoyancy is significantly smaller than one in 4 countries: the Netherlands, Israel, Slovak Republic and Hungary.
  - In 16 countries the coefficient is not significantly different from one.
  - In the remaining 14 countries, the coefficient significantly exceeds one.
- Cross-country variation:
  - Range observed: 0.75 for Slovak Rep. to 1.3 for Japan.
  - Standard deviation across countries: 0.11.

### Country-level buoyancy: long-run, short-run, and speed of adjustment
- Long-run buoyancy (country-level):
  - Range: 0.75 (Slovak Rep.) to 1.3 (Japan).
  - Mean: 1.032
  - Median: 1.048
  - St deviation: 0.110
  - In about half of the 34 OECD countries long-run buoyancy is not significantly different from one; in 14 countries it exceeds one.
- Short-run buoyancy (country-level):
  - Average short-run buoyancy: 1.01
  - Distribution:
    - Significantly smaller than one in 4 countries: Austria, Switzerland, Hungary and Estonia.
    - Significantly exceeds one in 5 countries: France, Japan, Spain, Australia and Chile.
    - Not significantly different from one in 25 countries.
  - Greater cross-country variation than long-run buoyancy: standard deviation of almost 0.2 (table reports 0.275 in one column and 0.196 in another).
  - Lowest short-run buoyancy: 0.43 in Hungary.
  - Highest short-run buoyancies: exceed 1.5 in Japan and Chile.
  - Comparison count: 18 countries have short-term buoyancy larger than long-run buoyancy; 16 have it smaller.
- Speed of adjustment (error-correction term):
  - Estimates are negative for all countries and statistically significant for most, indicating convergence to long-term relationship.
  - Range: from a low 7.8 percent in Sweden to over 75 percent in the Slovak and Czech Republics.
- Selected country examples:
  - Canada: Long-run 0.96***; Short-run 1.01***; Speed -0.10; Number of obs. 47.
  - France: Long-run 1.12***; Short-run 1.20***; Speed -0.16*; Number of obs. 47.
  - Japan: Long-run 1.29***; Short-run 1.54***; Speed -0.25***; Number of obs. 45.
  - Slovak Rep.: Long-run 0.75***; Short-run 0.84***; Speed -0.77***; Number of obs. 17.
  - Czech Rep.: Long-run 1.05***; Short-run 1.28***; Speed -0.78***; Number of obs. 17.
  - Hungary: Long-run 0.87***; Short-run 0.44***; Speed -0.59***; Number of obs. 21.

### Panel estimates and period comparison (balanced OECD panel, PMG estimator)
- Full period 1965-2012:
  - Long-run buoyancy: 1.06*** (standard error (0.00))
  - Short-run buoyancy: 1.04*** (s.e. (0.05))
  - Speed of adjustment: -0.17*** (s.e. (0.03))
- Subperiod 1965-1988:
  - Long-run buoyancy: 1.09*** (s.e. (0.00))
  - Short-run buoyancy: 0.93*** (s.e. (0.06))
  - Speed of adjustment: -0.33*** (s.e. (0.05))
- Subperiod 1989-2012:
  - Long-run buoyancy: 1.02*** (s.e. (0.03))
  - Short-run buoyancy: 1.17*** (s.e. (0.10))
  - Speed of adjustment: -0.39*** (s.e. (0.04))
- Interpretation:
  - On average long-run buoyancy is 1.06 and significantly higher than one for the full period.
  - Long-run buoyancy has declined since the late 1980s; after 1989 the coefficient is no longer significantly higher than one.
  - Short-run buoyancy rose markedly since the late 1980s: below one before the late 1980s (0.93***), above one after (1.17***); the difference between periods is statistically significant, implying a strengthened stabilization function in the more recent period.
- Panel coverage note: balanced panel excludes Chile, Czech Republic, Estonia, Hungary, Iceland, Israel, Korea, Mexico, the Netherlands, Poland, Slovakia and Slovenia; Australia and New-Zealand do not collect SSC.

### Buoyancy by tax revenue component (PMG panel estimates)
- Long-run buoyancy (with standard errors):
  - PIT: 0.97*** (0.01)
  - SSC: 1.05*** (0.01)
  - CIT: 1.26*** (0.02)
  - GST: 0.98*** (0.01)
  - Excises: 0.77*** (0.01)
  - Immovable Property: 0.71*** (0.02)
- Short-run buoyancy (with s.e.):
  - PIT: 1.10*** (0.09)
  - SSC: 0.75*** (0.07)
  - CIT: 1.96*** (0.17)
  - GST: 0.92*** (0.06)
  - Excises: 0.51*** (0.10)
  - Immovable Property: 0.05 (0.15)
- Speed of adjustment (with s.e.):
  - PIT: -0.20*** (0.04)
  - SSC: -0.20*** (0.03)
  - CIT: -0.30*** (0.03)
  - GST: -0.23*** (0.04)
  - Excises: -0.20*** (0.04)
  - Immovable Property: -0.15*** (0.03)
- Key takeaways:
  - CIT: highest long-run buoyancy (1.26***) and highest short-run buoyancy (1.96***), making it the best automatic stabilizer among tax categories.
  - PIT: long-run buoyancy below one (0.97***), short-run buoyancy above one (1.10***), suggesting it can act as an automatic stabilizer in the short run but long-run responsiveness is limited.
  - SSC: long-run buoyancy slightly above one (1.05***), but short-run buoyancy significantly smaller than one (0.75***), possibly due to regressive structure and contribution caps.
  - Excises and property taxes: lowest long-run buoyancy (0.77*** and 0.71*** respectively); short-run buoyancy weak for excises (0.51***) and not significantly different from zero for property taxes (0.05).

### Changes over time by tax component (1965-1988 vs 1989-2012, balanced panel)
- PIT:
  - Long-run buoyancy: 1.17*** → 0.82***
  - Short-run buoyancy: 1.07*** → 0.94***
  - Interpretation: decline consistent with reduced progressivity since late 1980s.
- SSC:
  - Long-run buoyancy: 1.23*** → 0.95***
  - Short-run buoyancy: 0.83*** → 0.58***
- CIT:
  - Long-run buoyancy: 1.24*** → 1.41***
  - Short-run buoyancy: 1.29*** → 3.58***
  - Interpretation: rise may reflect increase in capital income share over past decades.
- GST:
  - Long-run buoyancy: 1.30*** → 1.13***
  - Short-run buoyancy: 0.93*** → 1.28***
  - Interpretation: short-run increases may reflect procyclical consumption of luxuries.
- Excises:
  - Long-run buoyancy: 0.79*** → 0.52***
  - Short-run buoyancy: 0.43 → 0.36***
  - Interpretation: decline may be due to lack of indexation of excise rates.
- Property tax:
  - Long-run buoyancy: 0.66*** → 0.94***
  - Short-run buoyancy: around zero or negative in subperiods.

### Robustness checks: tax rates, inflation, and asymmetric responses
- Controlling for tax rates (selected results):
  - PIT (1989-2012):
    - Without control: Long-run 0.83*** (0.01)
    - Controlling for tax rate: Long-run 0.93*** (0.02)
    - Short-run: 0.95*** → 1.04***
    - Interpretation: controlling for top PIT rate increases estimated buoyancy, implying policy changes correlated with GDP may bias naive estimates.
  - CIT (1980-2012):
    - Long-run: 1.52*** (0.03) → 1.61*** (0.04) when controlling for tax rate
    - Short-run: 2.77*** → 2.89***
    - Interpretation: CIT results robust to inclusion of tax rate.
- Controlling for inflation:
  - Not controlling: Long-run 1.06*** (0.00); Short-run 1.04*** (0.05); Speed -0.17*** (0.03).
  - Controlling for inflation:
    - Long-run buoyancy: 0.79*** (0.02)
    - Short-run buoyancy: 0.89*** (0.08)
    - Long-run price effect: 0.37*** (0.03)
    - Short-run price effect: 0.20** (0.08)
    - Speed: -0.18*** (0.03)
  - Interpretation: inflation has a significant positive effect in both short- and long-run; buoyancy in real terms is smaller than in nominal terms and significantly so in the long-run.
- Asymmetric short-term buoyancy (growth vs contraction years):
  - Total revenue short-run buoyancy:
    - Growth years: 1.01*** (0.03)
    - Contraction years: 1.19*** (0.07)
  - Selected tax-type examples (short-run buoyancy):
    - SSC: growth 0.83*** (0.09); contraction 1.05*** (0.23)
    - CIT: growth 1.79*** (0.20); contraction 1.31*** (0.28)
    - Excises: growth 0.55*** (0.14); contraction 0.80*** (0.18)
    - Property tax: growth 0.14 (0.20); contraction 1.86** (0.83)
  - Caveat: contraction-year estimates have large standard errors due to fewer observations (142 observations for negative growth vs 1632 for positive growth) and should be interpreted with caution.

### Conclusions and suggested directions for further research
- Main conclusions:
  - Long-run buoyancy is close to one or slightly above one for many OECD countries; long-run buoyancy has declined since the late 1980s.
  - Short-run buoyancy is close to one for the majority of countries but has increased since the late 1980s.
  - The corporate income tax is the most effective automatic stabilizer (highest short-run buoyancy); excises and property taxes show weak short-run and long-run buoyancy.
  - Tax systems appear to operate better as automatic stabilizers during recessions than during expansions for total revenue and several tax categories.
- Three proposed directions to expand analysis:
  - Assess buoyancy for non-OECD countries (despite shorter time series), as tax ratios differ and buoyancy may differ outside the OECD.
  - Expand analysis to tax elasticities (which correct for changes in policy parameters), requiring detailed information on underlying tax reforms and revenue impacts.
  - Conduct a more thorough analysis of determinants of tax buoyancy to explain cross-country and over-time differences.

*Source: _wp14110*

### 1. Buoyancy of Total Tax Revenue by Country ......................................................................10

### 1. Buoyancy of Total Tax Revenue by Country ......................................................................10

### Context and key questions
- Many OECD countries face significant fiscal challenges after the crisis, with fiscal deficits having "skyrocketed" and ageing populations putting further pressure on public finances.
- Tax buoyancy measures how tax revenues vary with changes in GDP: a buoyancy of one implies an extra percent of GDP raises tax revenue by 1 percent (keeping the tax-to-GDP ratio unchanged); buoyancy > 1 raises revenue by more than GDP and can reduce the deficit ratio.
- Short-run buoyancy relates to fiscal stabilization (automatic stabilizers); long-run buoyancy matters for fiscal sustainability.
- The paper asks whether observed buoyancy > 1 during some years holds over longer periods, for all countries, and for all taxes.

### Methodology: specification and estimation
- Model: panel autoregressive distributed lag, transformed into an Error Correction Model (ECM) with p=1 and q=1 selected by the Bayesian information criterion.
- ECM parameters of interest:
  - θi,t=1 (short-term parameter) measures the instantaneous effect of a change in GDP on tax revenue (short-term buoyancy).
  - βi (long-run buoyancy) measures the long-run relationship between taxes and GDP.
  - λi (speed of adjustment) measures how fast buoyancy converges to its long-run equilibrium value.
- Estimation approaches:
  - Aggregate revenue: estimated per country using the mean group (MG) estimator; simple average of country estimates yields an OECD average.
  - Revenue by tax category and by period: estimated using panel regressions with the pooled mean group (PMG) estimator, which constrains long-run coefficients to be equal across groups while allowing intercepts and short-run coefficients to differ.

### Theoretical expectations
- Benchmark long-run buoyancy expected to be one.
- Long-run buoyancy likely larger for more progressive taxes (e.g., PIT) and smaller for regressive taxes (e.g., SSC, GST).
- VAT/GST buoyancy may exceed one if luxury goods (income elasticity > 1) are largely taxed at standard rates while necessities face reduced rates.
- Structural changes (e.g., decline in labor-income share) can reduce SSC buoyancy and increase CIT buoyancy.
- Excise taxes not indexed to GDP growth can yield buoyancy < 1.
- Short-run buoyancy differs by tax:
  - CIT expected to have relatively high short-run buoyancy (good automatic stabilizer).
  - Property taxes may have particularly small short-run buoyancy if local governments stabilize them counter-cyclically.
  - PIT and SSC may have short-run buoyancy below one in countries with rigid wages and tight employment protection.
  - GST short-run buoyancy can be below one due to consumption smoothing.
  - Recessions can reduce compliance and lower short-term buoyancy.

### Data
- Annual tax revenue data for 34 OECD countries, 1965–2012, from OECD revenue statistics.
- For countries joining OECD after 1965, data begin at accession; "The first year of data for the whole sample is 1995."
- Number of years per country varies from 16 for Estonia to 47 for the oldest members.
- GDP data from IMF sources.
- In 2012, the average tax-to-GDP ratio in the OECD is 34 percent.
- Revenue categories used: PIT, CIT, SSC, GST, Excises, and Property taxes.
- Observed composition dynamics (OECD average, balanced panel since 1965): steady increase in GST and SSC shares; PIT rose until late 1980s then started to fall. PIT and SSC are the two main sources of revenues.

### Results — Buoyancy of Total Tax Revenue (aggregate, country-by-country)
- Country-level estimation of Equation (2) for total tax revenue across 34 OECD countries yields:
  - An average long-run buoyancy of 1.03.
  - A median long-run buoyancy of 1.05.
- Significance across countries:
  - Long-run buoyancy is significantly smaller than one in 4 countries: the Netherlands, Israel, Slovak Republic and Hungary.
  - In 16 countries the coefficient is not significantly different from one.
  - In the remaining 14 countries, the coefficient significantly exceeds one.
- Cross-country variation in long-run buoyancy ranges (specific extremes are discussed in the full table in the source).

### Estimation approach implications and robustness
- MG estimator for country-level aggregate revenue imposes no common slope restrictions but may be inconsistent with limited data.
- PMG estimator for panel tax-category equations assumes common long-run buoyancy across countries but allows heterogeneity in short-run dynamics and error variances.
- The ECM framework permits simultaneous estimation of short-run effects, long-term relationship, and speed of adjustment, enabling decomposition of buoyancy into short-run and long-run components.

*Source: _wp14110 - 1. Buoyancy of Total Tax Revenue by Country ......................................................................10*

### 0.75 for Slovak Republic to 1.3 for Japan, with a standard deviation across countries of 0.11.

### _wp14110 - 0.75 for Slovak Republic to 1.3 for Japan, with a standard deviation across countries of 0.11.

### Country-level buoyancy: long-run, short-run, and speed of adjustment
- Long-run buoyancy:
  - Range observed: 0.75 for Slovak Rep. to 1.3 for Japan (table highlights country estimates).
  - Mean: 1.032
  - Median: 1.048
  - St deviation: 0.110
  - In about half of the 34 OECD countries long-run buoyancy is not significantly different from one; in 14 countries it exceeds one (implying GDP growth improved fiscal sustainability).
- Short-run buoyancy:
  - Average short-run buoyancy: 1.01
  - Distribution: in only four countries it is significantly smaller than one (Austria, Switzerland, Hungary and Estonia); in only five countries it significantly exceeds one (France, Japan, Spain, Australia and Chile); in 25 countries it is not significantly different from one.
  - Greater cross-country variation than long-run buoyancy: standard deviation of almost 0.2 (table reports 0.275 in one column and 0.196 in another; table-level St deviation entries: 0.275 and 0.196).
  - Lowest short-run buoyancy: 0.43 in Hungary.
  - Highest short-run buoyancies: exceed 1.5 in Japan and Chile.
  - Count comparison: 18 countries have short-term buoyancy larger than long-run buoyancy; 16 have it smaller.
- Speed of adjustment (error-correction term):
  - Estimates are negative for all countries and statistically significant for most, indicating convergence to long-term relationship.
  - Range: from a low 7.8 percent in Sweden to over 75 percent in the Slovak and Czech Republics.
- Selected country examples from Table 1 (preserve reported estimates where shown):
  - Canada: Long-run 0.96***; Short-run 1.01***; Speed -0.10; Number of obs. 47.
  - France: Long-run 1.12***; Short-run 1.20***; Speed -0.16*; Number of obs. 47.
  - Japan: Long-run 1.29***; Short-run 1.54***; Speed -0.25***; Number of obs. 45.
  - Slovak Rep.: Long-run 0.75***; Short-run 0.84***; Speed -0.77***; Number of obs. 17.
  - Czech Rep.: Long-run 1.05***; Short-run 1.28***; Speed -0.78***; Number of obs. 17.
  - Hungary: Long-run 0.87***; Short-run 0.44***; Speed -0.59***; Number of obs. 21.

### Panel estimates and period comparison (Table 2)
- Panel (PMG estimator) for full balanced OECD panel:
  - 1965-2012:
    - Long-run buoyancy 1.06*** (standard error (0.00) shown)
    - Short-run buoyancy 1.04*** (s.e. (0.05))
    - Speed of adjustment -0.17*** (s.e. (0.03))
  - 1965-1988:
    - Long-run buoyancy 1.09*** (s.e. (0.00))
    - Short-run buoyancy 0.93*** (s.e. (0.06))
    - Speed of adjustment -0.33*** (s.e. (0.05))
  - 1989-2012:
    - Long-run buoyancy 1.02*** (s.e. (0.03))
    - Short-run buoyancy 1.17*** (s.e. (0.10))
    - Speed of adjustment -0.39*** (s.e. (0.04))
- Interpretation:
  - On average long-run buoyancy is 1.06 and significantly higher than one for the full period.
  - Long-run buoyancy has declined since the late 1980s; after 1989 the coefficient is no longer significantly higher than one.
  - Short-run buoyancy rose markedly since the late 1980s: below one before the late 1980s (0.93***), above one after (1.17***); the difference between periods is statistically significant, implying a strengthened stabilization function in the more recent period.
  - Results control for balanced panel coverage; exclusions listed in source note (panel excludes Chile, Czech Republic, Estonia, Hungary, Iceland, Israel, Korea, Mexico, the Netherlands, Poland, Slovakia and Slovenia; Australia and New-Zealand do not collect SSC).

### Buoyancy by tax revenue component (Tables 3 and 4)
- Panel estimates for six tax categories (Table 3):
  - Long-run buoyancy (with standard errors in parentheses):
    - PIT: 0.97*** (0.01)
    - SSC: 1.05*** (0.01)
    - CIT: 1.26*** (0.02)
    - GST: 0.98*** (0.01)
    - Excises: 0.77*** (0.01)
    - Immovable Property: 0.71*** (0.02)
  - Short-run buoyancy (with s.e.):
    - PIT: 1.10*** (0.09)
    - SSC: 0.75*** (0.07)
    - CIT: 1.96*** (0.17)
    - GST: 0.92*** (0.06)
    - Excises: 0.51*** (0.10)
    - Immovable Property: 0.05 (0.15)
  - Speed of adjustment (with s.e.):
    - PIT: -0.20*** (0.04)
    - SSC: -0.20*** (0.03)
    - CIT: -0.30*** (0.03)
    - GST: -0.23*** (0.04)
    - Excises: -0.20*** (0.04)
    - Immovable Property: -0.15*** (0.03)
- Key takeaways:
  - CIT: highest long-run buoyancy (1.26***) and highest short-run buoyancy (1.96***), making it the best automatic stabilizer among tax categories.
  - PIT: long-run buoyancy below one (0.97***), short-run buoyancy above one (1.10***), suggesting it can act as an automatic stabilizer in the short run but long-run responsiveness is limited.
  - SSC: long-run buoyancy slightly above one (1.05***), but short-run buoyancy significantly smaller than one (0.75***), possibly due to regressive structure and contribution caps.
  - Excises and property taxes: lowest long-run buoyancy (0.77*** and 0.71*** respectively); short-run buoyancy weak for excises (0.51***) and not significantly different from zero for property taxes (0.05).
- Changes over time (Table 4, two periods 1965-1988 vs 1989-2012, balanced panel):
  - PIT long-run buoyancy fell from 1.17*** to 0.82***; short-run buoyancy fell from 1.07*** to 0.94***.
  - SSC long-run buoyancy fell from 1.23*** to 0.95***; short-run buoyancy fell from 0.83*** to 0.58***.
  - CIT long-run buoyancy rose from 1.24*** to 1.41***; short-run buoyancy rose from 1.29*** to 3.58***.
  - GST long-run buoyancy: 1.30*** (1965-1988) to 1.13*** (1989-2012); short-run buoyancy: 0.93*** to 1.28***.
  - Excises long-run buoyancy declined (0.79*** to 0.52***); short-run buoyancy remained below one (0.43 to 0.36***).
  - Property tax long-run buoyancy increased (0.66*** to 0.94***); short-run buoyancy around zero or negative in subperiods.
- Interpretation of component changes:
  - Decline in PIT buoyancy consistent with reduced progressivity since late 1980s.
  - CIT high buoyancy may reflect increase in capital income share over past decades.
  - GST short-run increases may reflect procyclical consumption of luxuries rather than consumption smoothing.
  - Excises low buoyancy may be due to lack of indexation of excise rates.

### Robustness checks: tax rates, inflation, and asymmetric responses
- Controlling for tax rates (Table 5):
  - PIT (1989-2012): without control long-run 0.83*** (0.01); controlling for tax rate long-run 0.93*** (0.02). Short-run buoyancy: 0.95*** → 1.04***.
  - CIT (1980-2012): long-run 1.52*** (0.03) → 1.61*** (0.04) when controlling for tax rate; short-run 2.77*** → 2.89***.
  - Interpretation: controlling for top PIT rate increases estimated buoyancy for PIT in 1989-2012, implying policy changes correlated with GDP may bias naive buoyancy estimates; CIT results robust to inclusion of tax rate.
- Controlling for inflation (Table 6):
  - Not controlling: Long-run buoyancy 1.06*** (0.00); Short-run 1.04*** (0.05); Speed -0.17*** (0.03).
  - Controlling for inflation: Long-run buoyancy 0.79*** (0.02); Short-run 0.89*** (0.08); Long-run price effect 0.37*** (0.03); Short-run price effect 0.20** (0.08); Speed -0.18*** (0.03).
  - Interpretation: inflation has a significant positive effect in both short- and long-run; buoyancy in real terms is smaller than in nominal terms and significantly so in the long-run.
- Asymmetric short-term buoyancy (Table 7):
  - Total revenue short-run buoyancy:
    - Growth years: 1.01*** (0.03)
    - Contraction years: 1.19*** (0.07)
  - For revenue types, short-run buoyancy is generally larger during contractions than during growth (notable examples):
    - SSC: 0.83*** (0.09) during growth; 1.05*** (0.23) during contraction.
    - CIT: 1.79*** (0.20) during growth; 1.31*** (0.28) during contraction.
    - Excises: 0.55*** (0.14) during growth; 0.80*** (0.18) during contraction.
    - Property tax: 0.14 (0.20) during growth; 1.86** (0.83) during contraction.
  - Caveat: results for contractions have large standard errors due to fewer observations (142 observations for negative growth vs 1632 for positive growth) and should be interpreted with caution.

### Conclusions and suggested directions for further research
- Main conclusions:
  - Long-run buoyancy is close to one or slightly above one for many OECD countries; long-run buoyancy has declined since the late 1980s.
  - Short-run buoyancy is close to one for the majority of countries but has increased since the late 1980s.
  - The corporate income tax is the most effective automatic stabilizer (highest short-run buoyancy); excises and property taxes show weak short-run and long-run buoyancy.
  - Tax systems appear to operate better as automatic stabilizers during recessions than during expansions for total revenue and several tax categories.
- Three proposed directions to expand analysis:
  - Assess buoyancy for non-OECD countries (despite shorter time series), as tax ratios differ and buoyancy may differ outside the OECD.
  - Expand analysis to tax elasticities (which correct for changes in policy parameters), requiring detailed information on underlying tax reforms and revenue impacts.
  - Conduct a more thorough analysis of determinants of tax buoyancy to explain cross-country and over-time differences.

*Source: IMF working paper content (excerpts from the provided PDF content).*

### References

### _wp14110 - References

### Tax base elasticities and tax buoyancy
- Bruce, D., W.F. Fox and M.H. Tuttle, 2006, Tax Base Elasticities: A Multi-State Analysis of Long-Run and Short-Run Dynamics, Southern Economic Journal 73(2), 315-41.
- Cotton, J.J., 2012, The Buoyancy and Elasticity of Non-Oil Tax Revenues in Trinidad and Tobago(1990-2009), Central Bank of Trinidad and Tobago Working Paper.
- Poghosyan, T., 2011, Assessing the Variability of Tax Elasticities in Lithuania, IMF Working paper 11/270.
- Upender, M., 2008, Degree Of Tax Buoyancy In India : An Empirical Study, International Journal of Applied Econometrics and Quantitative Studies 5(2), 59-70.
- Sobel, R.S. and R.G. Holcombe, 1996, Measuring Growth and Variability of Tax Bases Over the Business Cycle, National Tax Journal 49(4), 535-52.

### Property taxation and revenue potential
- Norregaard, J., 2013, Taxing Immovable Property: Revenue Potential and Implementation Challenges, IMF Working Paper 13/129.

### Methodology for panel estimation and cyclically-adjusted balances
- Pesaran, M.H., Y. Shin and R.P. Smith, 1999, Pooled Mean Group Estimation of Dynamic Heterogeneous Panels, Journal of the American Statistical Association 94(446), 621-634.
- Girouard, N. and C. André, 2005, Measuring Cyclically-adjusted Budget Balances for OECD Countries. OECD Economics Department Working Papers, Volume 434.

### Income distribution and wage shares
- Stockhammer, E., 2013, Why have wage shares fallen? A panel analysis of the determinants of functional income distribution, ILO Conditions of Work and Employment Series no. 35.

*Source: _wp14110 - References*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14110.pdf_