## 1. Incidence of Total Taxes

## Source details

**Canonical URL:** [1. Incidence of Total Taxes](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10112.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10112.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10112.pdf.json)

---

### Introduction and scope
- Central America comprises Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, Panama and the Dominican Republic.
- Purpose: survey existing tax and expenditure incidence studies for the region, assemble underlying data in a comparative framework to assess combined distributional impact of taxation and social spending in Central America.
- Main summary finding: the overall distributional effect of taxation in the region is small, while the redistributive impact of social spending is much larger, leading to a progressive net effect in all countries of the region.
- Policy-relevant conclusion: raising tax revenues and devoting the proceeds to social spending would unambiguously improve the income of the poorest households.

### Key limitations of the analysis
- Focus restricted to taxation and social spending; other spending components and indirect macroeconomic effects not addressed.
- Static incidence treatment (no analysis of education effects on future earnings).
- Interaction effects (behavioral responses) not examined.
- Efficiency, effectiveness, and administrative simplicity not assessed.
- Reliance on existing incidence studies limits cross-country comparability and timeliness; most underlying studies use data for 2000 (2003 for Panama and 2004 for Guatemala).

### Tax systems: structure and evolution
- In 2003, average central government tax burden in the region was around 12.5 percent of GDP.
- Tax collections increased in all countries except Guatemala between 2003 and 2008; in some cases by as much as 2 percentage points of GDP.
- Taxes account for most central government revenues across the region except Panama (taxes = 54 percent of central government revenue in 2008).
- Income taxes contribute on average about one-quarter of overall collection in Central America.
- Average share of trade taxes in total tax revenues is about 16 percent in Central America.
- VAT/sales/excise shares similar to OECD for taxes on goods and services.
- Other taxes (including property) play a relatively small role, except in Honduras.
- Structural shift 1995→2008: trade taxes fell from 20 percent to just over 10 percent of total tax revenue; VAT and income tax shares rose; excise share fell.
- Significant fraction of VAT revenue is collected at customs—VAT has often replaced import tariffs while leaving the object of taxation unchanged.

### Methodological considerations for incidence analysis
- Economic incidence assumptions (conventional in underlying studies):
  - Consumption taxes (VAT, sales, excise, import taxes) assumed fully shifted forward to consumers.
  - Export taxes paid by producers.
  - Personal income taxes paid by income recipients.
  - Corporate income taxes require assumptions about shifting (capital owners, workers, or consumers).
- Indicators used:
  - Tax progression: effective tax ratio per quantile; relative tax burden = effective tax rate of group divided by average tax rate.
  - Lorenz and concentration curves: compare pre-tax Lorenz curve with tax concentration curve; concentration curve below Lorenz curve → tax progressive over entire distribution (Lorenz dominance).

### Measures and indices of progressivity
- Quasi-Gini coefficient: Gini for a tax’s concentration curve. "The higher the quasi-Gini coefficient for a given tax, the more progressive it is."
- Kakwani (K) index: quasi-Gini (tax) minus Gini (pre-tax income). K > 0 → progressive; K < 0 → regressive.
- Reynolds-Smolensky (RS) index: pre-tax Gini minus quasi-Gini (post-tax income). RS sign consistent with K; RS captures redistributive effect incorporating tax pressure.
- Caveat: crossing of Lorenz and concentration curves (no Lorenz dominance) can make local indicators ambiguous; global indicators (quasi-Gini, K, RS) useful for complete ordering.

### Data and methodology notes for Central America
- Welfare measure: current total income used for consistency across countries.
- Data aggregation:
  - Honduras and Nicaragua incidence and social spending data based on quintiles.
  - Costa Rica and El Salvador decile-based data converted to quintiles.
- Coverage limited to central government taxes except:
  - Honduras includes municipal taxes.
  - Nicaragua includes taxes for the city of Managua.
- Exclusions: implicit taxes (price controls) and inflation tax excluded.
- Comparator countries: other Latin American countries, United States (federal taxes only), European Union.
- Note: larger number of groups raises several inequality indicators for a given underlying distribution.

### Distributional findings for Central America (income as welfare measure)
- Overall characterization:
  - "Tax systems in Central America are generally regressive."
  - Richer segments pay the bulk of taxes in absolute terms, but poorer segments pay more relative to income.
  - Negative Kakwani indices in many countries indicate regressivity; Guatemala and Panama ambiguous due to crossing curves and small positive Kakwani indices.
- Country highlights:
  - El Salvador: poorest quintile pays more than two and a half times as much taxes relative to income as the average citizen, and three and a half times what the richest quintile pays — high regressivity.
  - Costa Rica and Guatemala: relative burden fairly even; tax progression U-shaped—mildly regressive in lower quintiles and progressive at top quintiles (pattern favors middle classes).
  - Panama: U-shaped distribution with deeper trough; bottom quintile pays 28 percent more taxes than average household, top fifth pays 11 percent more.
  - Nicaragua: larger measured redistribution effect because regressivity combines with a relatively high ratio of taxes to household income.
- Consumption vs income:
  - Tax systems are "much less regressive or more progressive if consumption is used instead of income as a measure of welfare."
  - In El Salvador, Nicaragua, and Panama consumption is more evenly distributed than income; incidence measured relative to consumption is much less regressive.

### Magnitude of tax redistributive effects (selected Table 3 entries)
- Costa Rica (2000): Gini pre-tax income = 45.1; Quasi-Gini for taxes = 44.9; Kakwani = -0.2; Tax pressure = 20.8; Quasi-Gini for post-tax income = 45.1; RS = 0.0
- El Salvador (2000): Gini pre-tax income = 47.4; Quasi-Gini for taxes = 31.7; Kakwani = -15.7; Tax pressure = 8.1; Quasi-Gini for post-tax income = 48.8; RS = -1.4
- Guatemala (2004): Gini pre-tax income = 46.3; Quasi-Gini for taxes = 46.4; Kakwani = 0.1; Tax pressure = 17.3; Quasi-Gini for post-tax income = 46.3; RS = 0.0
- Honduras (2004): Gini pre-tax income = 47.2; Quasi-Gini for taxes = 40.8; Kakwani = -6.4; Tax pressure = 14.4; Quasi-Gini for post-tax income = 48.3; RS = -1.1
- Nicaragua (2000): Gini pre-tax income = 51.0; Quasi-Gini for taxes = 37.4; Kakwani = -13.6; Tax pressure = 27.5; Quasi-Gini for post-tax income = 56.2; RS = -5.2
- Panama (2003): Gini pre-tax income = 53.8; Quasi-Gini for taxes = 57.1; Kakwani = 3.3; Tax pressure = 6.4; Quasi-Gini for post-tax income = 53.6; RS = 0.2

### Progressivity of individual taxes (selected quasi-Gini (QG) and RS examples)
- VAT / Sales (QG; RS):
  - CRI: QG = 42.9; RS = -0.2
  - SLV: QG = 25.3; RS = -1.3
  - GTM: QG = 39.1; RS = 0.6
  - HND: QG = 31.3; RS = -1.2
  - NIC: QG = 27.8; RS = -2.9
  - PAN: QG = 50.5; RS = -0.1
- Excises (QG; RS):
  - CRI: QG = 49.3; RS = 0.1
  - SLV: QG = 37.9; RS = 0.0
  - GTM: QG = 47.9; RS = 0.0
  - HND: QG = 24.4; RS = -0.2
  - NIC: QG = 27.5; RS = -1.9
  - PAN: QG = 51.8; RS = 0.0
- International Trade (QG; RS):
  - CRI: QG = 34.4; RS = -0.2
  - SLV: QG = 31.8; RS = -0.2
  - GTM: QG = 58.3; RS = 0.4
  - HND: QG = 27.7; RS = -0.4
  - NIC: QG = 39.5; RS = -0.3
  - PAN: QG = 42.6; RS = -0.1
- Income taxes (QG; RS):
  - CRI: QG = 48.1; RS = 0.3
  - SLV: QG = 56.3; RS = 0.1
  - GTM: QG = 50.4; RS = 0.2
  - HND: QG = 61.6; RS = 0.7
  - NIC: QG = 58.6; RS = 0.4
  - PAN: QG = 73.9; RS = 0.4
- Notes:
  - Income taxes generally progressive but account on average for about one-fourth of total tax intake, so their redistributive contribution is small ("at or under 0.4 percentage points of the pre-tax Gini coefficient for all countries").
  - VAT and excises often regressive relative to income; incidence can be less regressive or progressive when measured relative to consumption.

### Distributional details by tax type (highlights)
- VAT/sales:
  - Clearly regressive when assessed relative to income in many Central American countries.
  - El Salvador: poorest 20 percent pay over three times more VAT relative to income than average household; poorest 20 percent pay five times as much VAT relative to income as richest 20 percent; consumption-to-income ratio = 177 percent (lowest quintile) vs 52 percent (highest quintile).
  - VAT regressivity is much lower when measured relative to consumption; VAT turns progressive in Panama, Guatemala, and Honduras when using consumption as welfare metric.
- Excise taxes:
  - Regressive in most Central American countries, except Costa Rica and Guatemala.
  - Honduras and Nicaragua: excises most regressive and sizable share → significant effect on overall distribution.
  - Regressivity driven by alcohol, tobacco, fuel taxes, which take larger share of poor households' consumption.
- International trade taxes (tariffs):
  - Highly regressive in most Central American countries as tariffs often applied to imported consumption goods consumed relatively more by poor households.
  - Guatemala appears an exception where tariffs may be concentrated on goods consumed by the rich.

### Social spending: trends, composition, and macro priorities
- Definition: social spending includes capital and current spending on education, health, social protection (social insurance and social assistance), housing, water and sewage, culture/sports/recreation.
- Regional trends:
  - Share of social spending in GDP in Central America was on average 11½ percent of GDP in 2004, an increase of 2¼ percent of GDP since 1995.
  - Average below Latin American average and well below Latin American median.
  - Social spending continued to increase in most countries except Costa Rica and Panama (already highest levels).
  - Public spending on education and health roughly similar to Latin American average; social protection spending much lower and varies significantly.
- Fiscal priority within public expenditures:
  - Share of social spending in total public expenditures increased.
  - Costa Rica: share at 68½ percent, 6 percentage points more than in 1994–95.
  - Honduras, Dominican Republic, Nicaragua: less than 40 percent of expenditure directed to social spending.

### Incidence and distributional impact of social spending
- Methodological notes:
  - Absolute incidence: share of total spending each income group receives.
  - Relative incidence: distribution of social spending relative to pre–fiscal income.
  - Quasi-Gini of spending and Kakwani index used; K < 0 denotes spending progressive relative to original income distribution.
- Aggregate findings:
  - Total public social spending is progressive in relative terms but not in absolute terms in several countries.
  - Costa Rica, Guatemala, Panama: positive quasi-Gini of spending (not progressive in absolute terms) but negative Kakwani indices (progressive in relative terms).
  - Magnitude of distributional impact (reductions in Gini due to social spending):
    - Costa Rica: reduction of 6 points.
    - Panama: reduction of 7.4 points.
    - El Salvador: reduction of 3.6 points.
  - International comparison:
    - Redistributive impact in Costa Rica and Panama comparable to some European countries (Italy, Spain, Portugal); exceeds impact in Andean countries.
    - Average quasi-Gini of –24.5 for social spending in EU-15 corresponds to 81 percent of social spending accruing to lowest three quintiles.
    - Lowest three quintiles receive about 70 percent of social spending in El Salvador; 60–65 percent in Costa Rica, Guatemala, Panama, Honduras; about 50 percent in Nicaragua.

### Social spending components and incidence
- Social protection (mainly pensions): pronouncedly regressive in Central America.
- Excluding social security:
  - Social spending is progressive in absolute terms in Costa Rica, El Salvador, and Panama.
  - Social spending progressive in relative terms in all Central American countries.
- Consistent with Lindert, Skoufias, and Shapiro (2006): social insurance programs often regressive in absolute terms across Latin America and the Caribbean.

### Key statistics from Table 5 (selected entries for social spending)
- Costa Rica (2000):
  - Pre-Spending Gini (Income) = 45.1
  - Quasi-Gini of Spending = 3.0
  - Kakwani Index = -42.1
  - Share of Social Spending 1/ = 18.2
  - Impact on Gini (RS Index) = 6.2
  - Post-Spending Gini = 38.9
- Guatemala (2004):
  - Pre-Spending Gini (Income) = 46.3
  - Quasi-Gini of Spending = 14.0
  - Kakwani Index = -32.3
  - Share of Social Spending 1/ = 6.3
  - Impact on Gini (RS Index) = 3.0
  - Post-Spending Gini = 43.3
- Panama (2003):
  - Pre-Spending Gini (Income) = 53.8
  - Quasi-Gini of Spending = 11.2
  - Kakwani Index = -42.7
  - Share of Social Spending 1/ = 17.4
  - Impact on Gini (RS Index) = 7.4
  - Post-Spending Gini = 47.0
- El Salvador (2000):
  - Pre-Spending Gini (Income) = 47.4
  - Quasi-Gini of Spending = -12.9
  - Kakwani Index = -60.3
  - Share of Social Spending 1/ = 5.3
  - Impact on Gini (RS Index) = 3.6
  - Post-Spending Gini = 43.8
- Honduras (2004):
  - Pre-Spending Gini (Income) = 47.2
  - Quasi-Gini of Spending = 0.7
  - Kakwani Index = -46.4
  - Share of Social Spending 1/ = 10.5
  - Impact on Gini (RS Index) = 3.2
  - Post-Spending Gini = 44.0
- Nicaragua (1998):
  - Pre-Spending Gini (Income) = 51.0
  - Quasi-Gini of Spending = 11.2
  - Kakwani Index = -39.8
  - Share of Social Spending 1/ = 8.6
  - Impact on Gini (RS Index) = 5.6
  - Post-Spending Gini = 45.5

### Incidence of education and health spending (selected findings)
- Education:
  - Total education spending progressive in relative terms in all countries; more so in El Salvador and Panama (progressive even in absolute terms).
  - Primary education: unambiguously pro-poor (strongly progressive in absolute terms) everywhere surveyed.
  - Secondary education: inverted U-shape (middle three quintiles benefit most); Guatemala exception with strong regressivity.
  - Tertiary education: regressive across the board; average ~25 percent of tertiary spending accrues to bottom 3 quintiles; Guatemala and Honduras tertiary spending regressive even in relative terms.
  - Reported QG and K for total education spending (selected):
    - CRI: QG = -1; K = -46
    - SLV: QG = -13; K = -60
    - GTM: QG = 3; K = -48
    - HND: QG = 5; K = -42
    - NIC: QG = 22; K = -29
    - PAN: QG = -4; K = -61
- Health:
  - Public health spending progressive in relative terms in all countries; progressive in absolute terms in four of seven countries.
  - Costa Rica and El Salvador: >25 percent of public health spending to poorest quintile; >70 percent to bottom three quintiles.
  - Guatemala, Honduras, Panama: neutral absolute incidence (~60 percent to bottom three quintiles).
  - Reported QG and K for health incidence (selected):
    - CRI: QG = -19; K = -64
    - DR: QG = -7; K = -50
    - SLV: QG = -13; K = -60
    - GTM: QG = 3; K = -48
    - HND: QG = 2; K = -45
    - NIC: QG = -2; K = -53
    - PAN: QG = 0; K = -56

### Net distributive effects of fiscal policy (taxes + social spending)
- Method: net direct distributional effect measured by RS index (positive → progressivity).
- Regional summary: net redistributive effect modestly progressive — taxation slightly regressive while social spending more strongly progressive, yielding positive RS index and fall in inequality.
- Reported Gini and RS values (Table 7, selected entries):
  - Costa Rica (2000): Pre-Fiscal Policy Gini = 45.1; Post-Taxation Gini = 45.1; Post-Social Spending Gini = 39.1; Post-Fiscal Policy Gini = 38.3; RS Index = 6.8
  - El Salvador (2000): 47.4; 48.8; 43.8; 45.8; RS Index = 1.6
  - Guatemala (2004): 46.3; 46.3; 43.2; 42.6; RS Index = 3.7
  - Honduras (2004): 47.2; 48.3; 44.0; 44.5; RS Index = 2.7
  - Nicaragua (1998): 51.0; 56.2; 45.5; 48.0; RS Index = 3.1
  - Panama (2003): 53.8; 53.6; 46.4; 45.8; RS Index = 8.0
  - EU-15 (2001): 41.7; 39.2; 32.8; 29.1; RS Index = 12.5
- Cross-regional pattern:
  - Central America and Andean countries: regressive tax systems offset by progressive social spending → modestly progressive net effect.
  - Europe (EU comparators): tax systems on average progressive and social spending highly progressive → large progressive net effect.

### Impact of fiscal policy on pre-fiscal income by quintile (Table 8 selected percentage changes)
- Costa Rica (quintiles 1–5): 59.1; 18.4; 6.1; -1.8; -10.1
- El Salvador: 19.0; 2.0; -1.8; -4.5; -5.2
- Guatemala: 12.8; 2.2; -3.1; -8.1; -15.0
- Honduras: 19.4; 1.9; -3.8; -6.6; -10.6
- Nicaragua: 8.6; 0.8; -3.9; -11.5; -14.8
- Panama: 161.8; 51.9; 22.8; 10.6; -3.0
- EU-15: 92.2; 39.6; 7.9; -8.0; -20.4
- Denmark: 164.5; 31.3; -19.6; -33.2; -44.2
- Ireland: 52.5; 78.3; 12.9; -6.8; -21.9
- Sweden: 114.1; 33.9; -6.5; -17.8; -28.9

### Simulated tax-financed increase in social spending (Table 9 scenarios)
- Scenario design: increase social spending by 1 percent of GDP financed by equivalent tax increase; four permutations:
  1. Increase in all taxes proportional to current shares, proceeds distributed according to current incidence of social spending.
  2. Increase in all taxes proportional to current shares, proceeds distributed equally to everyone.
  3. Increase in VAT only, proceeds distributed according to current incidence.
  4. Increase in VAT only, proceeds distributed equally to everyone.
- Key reported changes in Gini (Quasi-Gini change = post-reform minus pre-reform, times 100) — selected results:
  - Simulation 1 (overall tax, current incidence) — selected rows (quintiles 1–5; column 6 overall):
    - Costa Rica: 3.3; 1.4; 0.6; -0.1; -0.9; -0.6
    - El Salvador: 4.4; 1.3; 0.4; -0.2; -0.5; -0.4
    - Guatemala: 3.7; 2.1; 1.1; 0.2; -1.1; -0.7
    - Nicaragua: 3.3; 2.1; 1.3; 0.1; -0.8; -0.6
    - Panama: 5.4; 2.7; 1.2; 0.4; -1.2; -0.8
  - Simulation 2 (overall tax, evenly channeled):
    - Costa Rica: 2.5; 1.1; 0.7; 0.0; -0.8; -0.5
    - El Salvador: 3.2; 0.9; 0.3; -0.1; -0.4; -0.3
    - Nicaragua: 6.1; 2.7; 1.2; 0.1; -1.1; -0.8
    - Panama: 5.5; 2.3; 1.2; 0.4; -1.1; -0.7
  - Simulation 3 (VAT only, current incidence):
    - Costa Rica: 3.3; 1.4; 0.5; -0.2; -0.8; -0.6
    - El Salvador: 3.9; 1.1; 0.3; -0.2; -0.4; -0.4
    - Nicaragua: 2.2; 1.4; 0.9; -0.2; -0.4; -0.4
    - Panama: 5.4; 2.5; 0.9; 0.1; -1.0; -0.7
  - Simulation 4 (VAT only, evenly channeled):
    - Costa Rica: 2.5; 1.1; 0.6; 0.0; -0.7; -0.5
    - El Salvador: 2.7; 0.7; 0.2; -0.1; -0.3; -0.2
    - Guatemala: 4.2; 1.7; 0.6; -0.1; -0.8; -0.6
    - Nicaragua: 5.1; 2.0; 0.7; -0.2; -0.7; -0.6
    - Panama: 5.4; 2.0; 0.8; 0.1; -0.9; -0.7
- Aggregate simulation findings:
  - A 1 percent of GDP increase in tax revenue devoted to social spending reduces the income Gini coefficient between 0.5 and 0.8 percentage points in Costa Rica, Guatemala, Nicaragua, and Panama, and between 0.2 and 0.4 percentage points in El Salvador and Honduras.
  - Improving targeting of social spending can considerably increase redistribution (example: Nicaragua’s impact on the poorest quintile would double if regressivity were improved to at least a flat distribution).
  - Redistribution not much affected by whether financing comes from VAT or other taxes, even if VAT is regressive in isolation.

### Conclusions and policy recommendations
- Core conclusion: taxation has limited redistributive potential compared with social spending; well-targeted social spending is the more powerful tool for redistribution in Central America.
- Policy implications and recommendations:
  - Focus tax policy on raising revenue efficiently to expand resources available for redistribution through social spending.
  - Consider broadening the tax base—even if eliminating progressive exemptions—to increase resources for social spending and enhance overall progressivity.
  - Retain equity considerations in tax design; poorly designed exemptions can favor richer segments and harm equity.
  - Increase the absolute level of public social spending in the region; current levels generally low relative to GDP and total public spending.
  - Improve targeting of social spending:
    - Sustain progressive health and primary education spending.
    - Address regressive pensions and tertiary education by improving access and coverage to enhance impact on the poor.
    - Expand well-targeted social assistance (e.g., conditional cash transfers) to reduce poverty, including over the long run.
- Overall: combined taxation and well-targeted social spending can substantially improve incomes of the poor, even when the tax system alone is regressive.

### Appendix 1 — On what basis should the tax burden be measured?
- Income vs consumption trade-offs:
  - Current income widely used but volatile, subject to survey under-reporting (self-employment, capital income), poor capture of remittances, and insensitive to household composition.
  - Consumption is less volatile, better proxy for permanent income, and typically more evenly distributed—incidence studies using consumption often find greater progressivity.
- Empirical evidence in Central America:
  - Studies using consumption as the welfare metric find greater progressivity in El Salvador, Nicaragua, Panama, Guatemala, and Honduras.
- Limitations of consumption:
  - May mis-measure permanent income with bequest motives or precautionary savings.
  - Data constraints: many household surveys do not measure consumption; cross-country comparisons limited—income used in this paper due to that constraint.
- Appendix sources: Barreix, Roca, and Villela (2006); Auguste and Artana (2005); plus methodological literature cited.

*Source: IMF staff paper segment titled "1. Incidence of Total Taxes" (Central America tax and social spending incidence analysis).*

### 1. Incidence of Total Taxes ............................................................................................

### 1. Incidence of Total Taxes

### Introduction: purpose and scope
- Central America comprises Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua, Panama and the Dominican Republic.
- The paper surveys existing tax and expenditure studies for the countries in the region, assembling underlying data in a comparative framework to assess the combined distributional impact of taxation and social spending in Central America.
- Main summary finding: the overall distributional effect of taxation in the region is small, while the redistributive impact of social spending is much larger, leading to a progressive net effect in all countries of the region.
- Policy-relevant conclusion highlighted: raising tax revenues and devoting the proceeds to social spending would unambiguously improve the income of the poorest households.

### Key limitations of the analysis
- Focus is on taxation and social spending only; other components of spending and indirect effects of overall fiscal stance are not addressed.
- Incidence and distributional impact are treated in a static sense (for example, spending on education is not analyzed for effects on future earning potential).
- Interaction effects (taxation and public provision of social services and transfers affecting behavior) are not examined.
- Efficiency, effectiveness, and administrative simplicity of taxation and social spending are not assessed.
- Analysis relies on existing incidence studies, limiting cross-country comparability and timeliness; most underlying studies use data for 2000 (2003 for Panama and 2004 for Guatemala).

### Tax systems in Central America: structure and evolution
- In 2003, the average tax burden of the central governments in the region was around 12.5 percent of GDP, only marginally higher than its 1995 level.
- More recently, tax collections in Central America increased in all countries except Guatemala, in some cases by as much as 2 percentage points of GDP between 2003 and 2008 (see Appendix Table A.1).
- Tax revenue accounts for most central government revenues across the region; exception: Panama, where significant income from the Panama Canal drives down the share of taxes in central government revenue (just 54 percent in 2008).
- Income taxes contribute on average only about one-quarter of overall collection in Central America (and Latin America as a whole), compared with one-half in the OECD.
- The average share of trade taxes in total tax revenues is about 16 percent in Central America, compared with only 1 percent in the OECD.
- Taxes on goods and services (VAT, sales, and excise taxes) account for similar shares of total revenue in Central America and the OECD.
- Other taxes, including property taxes, play a relatively small role in Central America (with the exception of Honduras), Latin America, and, to a lesser extent, OECD countries.
- Structural shift observed: between 1995 and 2008 the share of trade taxes in total tax revenue fell from a regional average of 20 percent to just over 10 percent, with the decline largely made up by increases in the VAT and income tax shares, while the share of excise taxes has also fallen.
- Note: a significant fraction of VAT revenue throughout the region is collected at customs; in this sense, while the VAT has replaced import tariffs, the object of taxation remains unchanged.

### Distributional effects of taxation: methodological considerations
- Incidence analysis requires assumptions about economic incidence (who ultimately bears the burden), which can differ from legal incidence.
- Conventional assumptions used by the underlying studies:
  - Consumption taxes (VAT, sales, excise, and import taxes) are fully shifted forward to consumers.
  - Export taxes are paid by producers.
  - Personal income taxes are paid by the income recipients.
  - Corporate income taxes require more demanding assumptions (may be shifted backward to capital owners or workers or forward to consumers depending on mobility of capital).
- Indicators used to measure distributional impact:
  - Tax progression: measures the effective tax ratio (tax effectively paid relative to income) per quantile. The paper uses a normalized measure of tax progression—the relative tax burden—defined as the effective tax rate that each income group pays divided by the average tax rate for the population as a whole.
  - Lorenz and concentration curves: progression of a tax is represented by a concentration curve (cumulative tax paid per quantile of pre-tax income). Progressivity assessed by comparing the pre-tax Lorenz curve for income with the concentration curve for that tax; a tax is progressive over the entire distribution if the concentration curve lies consistently below the pre-tax Lorenz curve (Lorenz dominance).

### Empirical and policy takeaways (from the surveyed evidence)
- Empirical evidence for developed and developing countries suggests the overall effect of taxes on income distribution is generally limited, and relatively profound changes in tax structures have only a small distributional impact.
- In contrast, the distributional effects of public spending, especially of well-targeted social spending, can have substantial positive effects on equity and poverty reduction.
- Many analysts conclude tax policy should focus on efficiency issues, with the redistributive aim of fiscal policy accomplished via the expenditure side.
- Nonetheless, understanding the distributional effects of certain taxes and their determinants can help shape more equitable tax systems without necessarily sacrificing efficiency.
- Regional evidence and country-level incidence studies indicate that recent major tax reforms (for example, in Nicaragua and Guatemala as analyzed by earlier studies) had only small distributional impacts; two additional reforms (Guatemala 2006 and Nicaragua 2009) are noted but not assessed in the underlying studies.
- Given the relatively small redistributive impact of taxes and larger redistributive impact of social spending, policy options that combine increased revenue mobilization with allocation to social spending are highlighted as effective for improving incomes of the poorest households.

*Source: IMF staff paper segment titled "1. Incidence of Total Taxes" (Central America tax and social spending incidence analysis).*

### Conclusions on the distributional effects of taxation are sensitive to incidence assumptions. They must,

### Conclusions on the distributional effects of taxation are sensitive to incidence assumptions. They must,

### Measures and indices of tax progressivity
- Quasi-Gini coefficient: the Gini coefficient for a tax’s concentration curve. "The higher the quasi-Gini coefficient for a given tax, the more progressive it is."
- Kakwani (K) index: difference between the quasi-Gini coefficient for a given tax and the Gini coefficient for pre-tax income.
  - If K > 0, the tax is progressive.
  - If K < 0, the tax is regressive.
- Reynolds-Smolensky (RS) index: defined as the pre-tax Gini coefficient minus the quasi-Gini index for post-tax income. It measures how income inequality changes (in Gini points) as a result of the introduction of the tax.
  - The sign of RS is consistent with that of K (positive = progressive; negative = regressive).
  - RS captures the redistributive effect by incorporating the importance of tax revenues relative to the economy (tax pressure).

- Conceptual difference:
  - Concentration curve: plots cumulative shares of X (e.g., tax payments) with respect to the quantile distribution of Y (e.g., pre-tax income).
  - Lorenz curve: represents cumulative share of Y with respect to the quantile distribution of Y.
- Caveat: if pretax Lorenz and concentration curves cross (Lorenz dominance fails), local indicators (progression and concentration curves) may be ambiguous; global indicators (quasi-Gini, K, RS) are useful for complete ordering.

### Data and methodology notes for Central America incidence analysis
- Welfare measure: current total income used for consistency across countries.
- Data aggregation:
  - Honduras and Nicaragua incidence data, and social spending data in all countries, are based on quintiles of income.
  - Costa Rica and El Salvador incidence data originally based on deciles were converted to quintiles to enable comparisons.
- Coverage: underlying data limited to central government taxes except:
  - Honduras includes municipal taxes.
  - Nicaragua includes taxes for the city of Managua.
- Exclusions: implicit taxes (such as price controls) and the inflation tax are excluded from the incidence analysis.
- Comparator countries: other Latin American countries, the United States (federal taxes only), and the European Union selected for relevance and data comparability.
- Note on indicators: "For a given underlying distribution, the larger the number of groups, the higher the several indicators of inequality will be."

### Distributional findings for Central America (summarized)
- Overall characterization:
  - "Tax systems in Central America are generally regressive."
  - Richer segments pay the bulk of taxes (Table 2, Panel B), but the poor pay more taxes relative to income (Table 2, Panel C).
  - Negative Kakwani indices in many countries indicate regressivity; Guatemala and Panama show no unambiguous conclusion due to crossing curves and small positive Kakwani indices.
- Country-specific highlights:
  - El Salvador: the poorest quintile pays more than two and a half times as much taxes relative to their income as the average citizen, and three and a half times what the richest quintile pays—high regressivity.
  - Costa Rica and Guatemala: relative burden distributed fairly evenly; tax progression U-shaped—mildly regressive in lower quintiles and progressive at the top quintiles (pattern favors middle classes).
  - Panama: U-shaped distribution with a deeper trough; bottom quintile pays 28 percent more taxes than the average household, top fifth pays 11 percent more.
  - Nicaragua: measured redistribution effect somewhat larger because regressivity combines with a relatively high ratio of taxes to household income.
- Consumption vs income:
  - Tax systems are "much less regressive or more progressive if consumption is used instead of income as a measure of welfare."
  - In El Salvador, Nicaragua, and Panama consumption/expenditure is more evenly distributed than income (Appendix Table A.4); incidence measured relative to consumption is much less regressive.

### Magnitude of tax redistributive effects
- Redistributive impact is limited across the region: "Tax systems in the region, whether progressive or regressive, have a limited effect on the overall distribution of income."
- Determinants of redistributive impact:
  - Pre-tax income distribution.
  - Distribution of tax payments across income groups.
  - Tax pressure: ratio of total taxes considered in the incidence analysis to total income before taxes.
- Table 3 summary (selected values; RS = Reynolds-Smolensky Index; positive RS denotes progressivity):
  - Costa Rica (2000): Gini pre-tax income = 45.1; Quasi-Gini for taxes = 44.9; Kakwani = -0.2; Tax pressure = 20.8; Quasi-Gini for post-tax income = 45.1; RS = 0.0
  - El Salvador (2000): Gini pre-tax income = 47.4; Quasi-Gini for taxes = 31.7; Kakwani = -15.7; Tax pressure = 8.1; Quasi-Gini for post-tax income = 48.8; RS = -1.4
  - Guatemala (2004): Gini pre-tax income = 46.3; Quasi-Gini for taxes = 46.4; Kakwani = 0.1; Tax pressure = 17.3; Quasi-Gini for post-tax income = 46.3; RS = 0.0
  - Honduras (2004): Gini pre-tax income = 47.2; Quasi-Gini for taxes = 40.8; Kakwani = -6.4; Tax pressure = 14.4; Quasi-Gini for post-tax income = 48.3; RS = -1.1
  - Nicaragua (2000): Gini pre-tax income = 51.0; Quasi-Gini for taxes = 37.4; Kakwani = -13.6; Tax pressure = 27.5; Quasi-Gini for post-tax income = 56.2; RS = -5.2
  - Panama (2003): Gini pre-tax income = 53.8; Quasi-Gini for taxes = 57.1; Kakwani = 3.3; Tax pressure = 6.4; Quasi-Gini for post-tax income = 53.6; RS = 0.2
- Comparative perspective:
  - With exceptions (Guatemala and Panama), Central American taxation is generally more regressive and has lower quasi-Gini indices than Andean countries, the United States (federal), and the European Union.
  - Sweden and Denmark are noted exceptions where tax systems are regressive, yet fiscal policy overall is powerfully progressive.

### Progressivity of individual taxes
- Income taxes:
  - "Income taxes are generally progressive in Central America."
  - Contribution to overall redistributive impact is small because income taxes account on average for about one-fourth of total tax intake across Central America.
  - Redistributive effect of income taxes: "at or under 0.4 percentage points of the pre-tax Gini coefficient for all countries."
  - Panama: income taxes are strongly progressive and account for a greater share of total tax revenue, but low share in income still results in a small redistributive effect.
- VAT / Sales, Excises, International Trade, Income (Figure 2 panels):
  - VAT / Sales: example quasi-Gini and RS entries
    - Costa Rica (CRI): QG = 42.9; RS = -0.2
    - El Salvador (SLV): QG = 25.3; RS = -1.3
    - Guatemala (GTM): QG = 39.1; RS = 0.6
    - Honduras (HND): QG = 31.3; RS = -1.2
    - Nicaragua (NIC): QG = 27.8; RS = -2.9
    - Panama (PAN): QG = 50.5; RS = -0.1
  - Excises: example quasi-Gini and RS entries
    - CRI QG = 49.3; RS = 0.1
    - SLV QG = 37.9; RS = 0.0
    - GTM QG = 47.9; RS = 0.0
    - HND QG = 24.4; RS = -0.2
    - NIC QG = 27.5; RS = -1.9
    - PAN QG = 51.8; RS = 0.0
  - International Trade: example quasi-Gini and RS entries
    - CRI QG = 34.4; RS = -0.2
    - SLV QG = 31.8; RS = -0.2
    - GTM QG = 58.3; RS = 0.4
    - HND QG = 27.7; RS = -0.4
    - NIC QG = 39.5; RS = -0.3
    - PAN QG = 42.6; RS = -0.1
  - Income tax panel QG and RS examples
    - CRI QG = 48.1; RS = 0.3
    - SLV QG = 56.3; RS = 0.1
    - GTM QG = 50.4; RS = 0.2
    - HND QG = 61.6; RS = 0.7
    - NIC QG = 58.6; RS = 0.4
    - PAN QG = 73.9; RS = 0.4

### Overall synthesis
- Distributional conclusions depend critically on incidence assumptions and the choice of welfare measure (income vs consumption).
- Central American tax systems are commonly regressive when welfare is measured by income, with notable country heterogeneity (El Salvador, Honduras, Nicaragua most regressive; Costa Rica and Guatemala more evenly distributed; Panama mixed).
- Even where taxes are progressive, their redistributive impact is modest due to low tax pressure and limited revenue shares of progressive taxes (notably income taxes).
- Measuring welfare by consumption/expenditure reduces apparent regressivity in several countries.

*Source: IMF staff analysis and country incidence studies as summarized in the supplied PDF content.*

### Appendix Tables A.2 and A.3 also show that income taxes are much more progressive in

### _wp10112 - Appendix Tables A.2 and A.3 also show that income taxes are much more progressive in

### Tax incidence and distributional findings
- Income taxes:
  - Income taxes are much more progressive in Andean countries and the United States (except, in some cases, with respect to Panama and Honduras), and their distributional effect is much stronger in these comparator countries.
  - Income taxes without social security are negative (that is, after-tax income is higher) for the bottom two income quintiles in the United States because of earned income tax credits. (Footnote 32)

- Value-added taxes (VAT) and sales taxes:
  - VATs and sales taxes are clearly regressive in Central America when assessed relative to income (Appendix Table A.2 and Figure 2).
  - El Salvador:
    - The poorest 20 percent of the population pay over three times more VAT relative to their income than the average household.
    - The poorest 20 percent pay five times as much VAT relative to their income as the richest 20 percent.
    - Ratio of consumption to income: 177 percent for the lowest quintile and 52 percent for the highest quintile.
  - VAT regressivity has a tangible effect on overall income distribution in El Salvador, Honduras, and especially Nicaragua, as reflected in highly negative Reynolds-Smolensky indices (Appendix Table A.3).
  - Costa Rica:
    - Local and global indicators of VAT regressivity are much lower, possibly reflecting targeted exemptions (exclusion of a basic basket of goods and services consumed mostly by the poor).
  - Measurement note:
    - The regressivity of the VAT in Central American countries is much lower if measured relative to consumption—consumption is more evenly spread than income, yielding higher consumption-to-income ratios for the poorest groups.
    - If consumption is used instead of current income, the VAT becomes much less regressive in El Salvador and Nicaragua, and turns progressive in Panama, Guatemala, and Honduras (Appendix Table A.4 and cited studies).
    - In the Dominican Republic, the VAT is highly progressive relative to household expenditure: the effective tax rate paid by the richest quintile is twice as large as that paid by the poorest (Jenkins, Jenkins, and Kuo, 2006).
    - The VAT remaining regressive in El Salvador and Nicaragua even when measured relative to consumption suggests exemptions may disproportionately benefit the rich in these countries. (Footnote 35)
  - Cross-country evidence:
    - Some studies indicate VATs are regressive relative to income (e.g., Gemmel and Morrissey, 2005), while evidence exists of progressive VAT incidence in some African countries (Sahn and Younger, 1999; Muñoz and Cho, 2004). Appendix Table A.3 shows the VAT is progressive in Ecuador and Venezuela even when considered relative to income. (Footnote 33–34)

- Excise taxes:
  - Excise taxes are regressive in most Central American countries, except in Costa Rica and Guatemala.
  - In Honduras and Nicaragua, excise taxes are the most regressive tax and have a sizable share in total taxation, producing a palpable effect on overall income distribution as indicated by the Reynolds-Smolensky index (Appendix Tables A.2 and A.3).
  - The regressive incidence of excise taxes is driven mainly by taxes on alcohol, tobacco, and fuel, because consumption of these goods accounts for a larger share of the income of poorer households. (Footnote 36)
  - Panama: excise taxes are regressive as a whole, but much less so; some excises (cars, luxury items) are progressive. (Footnote 37)
  - Costa Rica: excises are fairly progressive where broader coverage includes luxury goods; however, a legal reform in 2001 may have reduced or eliminated progressivity. (Footnote 38)
  - International evidence: excise taxes are highly progressive in Bolivia but regressive in the United States (Appendix Tables A.2 and A.3).

- International trade taxes (import tariffs):
  - Mostly import tariffs in the region are highly regressive in most Central American countries (Appendix Tables A.2 and A.3).
  - This regressivity reflects higher tariffs on imported consumption goods that are also produced domestically (especially food and lightly processed manufactured goods), which are a larger share of poor households’ consumption baskets.
  - Guatemala appears to be an exception, suggesting imported goods subject to tariffs may be more prominent in the consumption patterns of the rich in that country. (Footnote 39)

### Social spending: trends and macro priorities
- Definition:
  - Social spending includes capital and current spending on education, health, social protection (social insurance and social assistance), housing, water and sewage, and culture, sports, and recreation.
- Regional trends:
  - The share of social spending in GDP in Central America was on average 11½ percent of GDP in 2004, an increase of 2¼ percent of GDP since 1995 (Table 4).
  - This average is slightly below the Latin American average but well below the Latin American median.
  - Social spending continued to increase in recent years in most countries, with the exception of Costa Rica and Panama, which already devote the highest amount of resources to social spending. (Appendix Table A.5)
  - The macroeconomic priority has increased for all components of social spending.
  - Public spending on education and health in the region is roughly similar to the Latin American average (and median).
  - Spending on social protection is much lower and varies significantly across Central America, reflecting large differences in pension and social assistance spending (including conditional cash transfer programs) (Figure 3).
- Fiscal priority within public expenditures:
  - The share of social spending in total public expenditures has increased (Figure 4).
  - Costa Rica: highest share at 68½ percent, 6 percentage points more than in 1994–95.
  - Honduras, the Dominican Republic, and Nicaragua direct less than 40 percent of expenditure to social spending.

### Incidence and distributional impact of social spending
- Methodological considerations:
  - Incidence analysis requires identifying actual beneficiaries; some beneficiaries can be identified directly (e.g., primary spending via primary school enrollment ratios), while others require proxies (e.g., social assistance to the disabled).
  - Distinction:
    - Absolute incidence: share of total spending each income group receives.
    - Relative incidence: distribution of social spending relative to pre–fiscal policy income.
  - Indicators:
    - Quasi-Gini coefficient of spending: Gini coefficient for concentration curve of spending; values lie between –1 and 1; a negative value denotes progressivity in absolute terms (concentration curve above 45-degree line).
    - Kakwani index (K): difference between the quasi-Gini coefficient of spending and the Gini coefficient of original income distribution; if K < 0, spending is progressive relative to the original income distribution.

- Aggregate findings for Central America:
  - Total public social spending is progressive in relative but not absolute terms.
  - Costa Rica, Guatemala, and Panama (countries with comprehensive incidence studies):
    - All have positive quasi-Gini coefficients of spending, meaning social spending is not progressive in absolute terms (Table 5, Panel A, column 2).
    - Social spending is more equally distributed than pre-spending income and thus progressive in relative terms as indicated by negative Kakwani indices (Table 5, Panel A, column 3).
  - Magnitude of distributional impact varies considerably:
    - Reductions in Gini coefficient due to social spending:
      - Costa Rica: reduction of 6 points.
      - Panama: reduction of 7.4 points.
      - El Salvador: reduction of 3.6 points (on par with Guatemala and Honduras).
    - High levels of social spending as a share of GDP partly explain Costa Rica and Panama’s stronger distributional impacts.
  - International comparison:
    - In absolute terms (reduction in pre-fiscal Gini), the redistributive impact in most progressive Central American countries (Costa Rica and Panama) is comparable to some European countries (e.g., Italy, Spain, Portugal) and exceeds that observed in Andean countries.
    - Because of high initial inequality, post–social spending Ginis in Central America (except for Costa Rica) remain above pre–social-spending Ginis of European countries (Table 5, column 6).
    - The average quasi-Gini coefficient of –24.5 for social spending in the EU-15 corresponds to 81 percent of social spending accruing to the lowest three quintiles. By contrast:
      - Lowest three quintiles receive about 70 percent of social spending in El Salvador;
      - 60–65 percent in Costa Rica, Guatemala, Panama, and Honduras;
      - About 50 percent in Nicaragua.

- Social spending components:
  - Public spending on social protection (mainly pensions) is pronouncedly regressive in Central America.
  - If social security is excluded, social spending is:
    - Progressive in absolute terms in Costa Rica, El Salvador, and Panama (strongly pro-poor).
    - Progressive in relative terms in all Central American countries (Tables 5 and 6).
  - These findings align with Lindert, Skoufias, and Shapiro (2006), which found that social insurance programs studied are regressive in absolute terms across Latin America and the Caribbean.

### Key statistics from Table 5 (selected entries)
- Costa Rica (2000):
  - Pre-Spending Gini (Income): 45.1
  - Quasi-Gini of Spending: 3.0
  - Kakwani Index: -42.1
  - Share of Social Spending 1/: 18.2
  - Impact on Gini (RS Index): 6.2
  - Post-Spending Gini: 38.9
- Guatemala (2004):
  - Pre-Spending Gini (Income): 46.3
  - Quasi-Gini of Spending: 14.0
  - Kakwani Index: -32.3
  - Share of Social Spending 1/: 6.3
  - Impact on Gini (RS Index): 3.0
  - Post-Spending Gini: 43.3
- Panama (2003):
  - Pre-Spending Gini (Income): 53.8
  - Quasi-Gini of Spending: 11.2
  - Kakwani Index: -42.7
  - Share of Social Spending 1/: 17.4
  - Impact on Gini (RS Index): 7.4
  - Post-Spending Gini: 47.0
- El Salvador (2000):
  - Pre-Spending Gini (Income): 47.4
  - Quasi-Gini of Spending: -12.9
  - Kakwani Index: -60.3
  - Share of Social Spending 1/: 5.3
  - Impact on Gini (RS Index): 3.6
  - Post-Spending Gini: 43.8
- Honduras (2004):
  - Pre-Spending Gini (Income): 47.2
  - Quasi-Gini of Spending: 0.7
  - Kakwani Index: -46.4
  - Share of Social Spending 1/: 10.5
  - Impact on Gini (RS Index): 3.2
  - Post-Spending Gini: 44.0
- Nicaragua (1998):
  - Pre-Spending Gini (Income): 51.0
  - Quasi-Gini of Spending: 11.2
  - Kakwani Index: -39.8
  - Share of Social Spending 1/: 8.6
  - Impact on Gini (RS Index): 5.6
  - Post-Spending Gini: 45.5

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10112.pdf*

### 2004. For Europe, 2001data.

### _wp10112 - 2004. For Europe, 2001data.

### Incidence of public spending on education
- Total education spending is progressive in relative terms in all of Central America, more so in El Salvador and Panama (progressive even in absolute terms) and less so in Nicaragua (the richest quintile receives about 35 percent of total public spending).
- Public spending by level:
  - Primary education: unambiguously pro-poor (strongly progressive in absolute terms) in all countries surveyed.
  - Secondary education: follows an inverted U-shape; highest share accrues to the middle three quintiles in most countries, except Guatemala where it exhibits strong regressivity.
  - Tertiary education: regressive across the board; average of only 25 percent of public spending on tertiary education accrues to the bottom 3 quintiles.
    - In Guatemala and Honduras, tertiary education spending is regressive even in relative terms (positive Kakwani index).
- Reported Quasi-Gini (QG) and Kakwani (K) indices for incidence of total education spending (selected entries):
  - CRI: QG -1; K -46
  - SLV: QG -13; K -60
  - GTM: QG 3; K -48
  - HND: QG 5; K -42
  - NIC: QG 22; K -29
  - PAN: QG -4; K -61
- For tertiary education QG and K (selected entries):
  - CRI: QG 43; K -2
  - DR: QG 36; K -7
  - SLV: QG 47; K -1
  - GTM: QG 69; K 18
  - HND: QG 57; K 9
  - PAN: QG 36; K -21

### Incidence of public spending on health
- Distribution of public spending on health is progressive in relative terms in all Central American countries surveyed, and progressive in absolute terms in four of seven countries.
- Costa Rica and El Salvador: more than 25 percent of total public health spending goes to the poorest quintile, and more than 70 percent to the bottom three quintiles.
- Nicaragua and the Dominican Republic: modest but pro-poor incidence.
- Guatemala, Honduras, and Panama: neutral absolute incidence, with about 60 percent of spending accruing to the bottom three quintiles (in proportion with their share of total income).
- Reported QG and K for health incidence (selected entries):
  - CRI: QG -19; K -64
  - DR: QG -7; K -50
  - SLV: QG -13; K -60
  - GTM: QG 3; K -48
  - HND: QG 2; K -45
  - NIC: QG -2; K -53
  - PAN: QG 0; K -56

### Net distributive effects of fiscal policy (taxation and social spending)
- Method: Net direct distributional effect measured by Reynolds-Smolensky (RS) index (positive values denote progressivity).
- Regional summary: For Central America the net redistributive effect of fiscal policy is modestly progressive; taxation tends to be slightly regressive while social spending is more strongly progressive, yielding a positive RS index and a fall in inequality.
- Reported Gini and RS index values (Table 7, selected entries):
  - Costa Rica (2000): Pre-Fiscal Policy Gini (Income) 45.1; Post-Taxation Gini 45.1; Post-Social Spending Gini 39.1; Post-Fiscal Policy Gini 38.3; RS Index 6.8
  - El Salvador (2000): 47.4; 48.8; 43.8; 45.8; RS Index 1.6
  - Guatemala (2004): 46.3; 46.3; 43.2; 42.6; RS Index 3.7
  - Honduras (2004): 47.2; 48.3; 44.0; 44.5; RS Index 2.7
  - Nicaragua (1998): 51.0; 56.2; 45.5; 48.0; RS Index 3.1
  - Panama (2003): 53.8; 53.6; 46.4; 45.8; RS Index 8.0
  - EU-15 (2001): 41.7; 39.2; 32.8; 29.1; RS Index 12.5
- Cross-regional pattern:
  - Central America and Andean countries: tax systems tend to be regressive but are offset by progressive social spending → modestly progressive net effect.
  - Europe (EU comparators): tax systems on average progressive and social spending highly progressive → large and strongly progressive net effect (post–fiscal policy quasi-Gini index 12.5 points lower than pre–fiscal policy Gini; income of poorest quintile almost doubles).

### Impact of fiscal policy on pre-fiscal policy income, by quintile (Table 8, selected entries)
- Reported percentage changes in pre-fiscal policy income by quintile due to fiscal policy (columns 1–5 correspond to quintiles 1–5, lowest to highest):
  - Costa Rica: 59.1; 18.4; 6.1; -1.8; -10.1
  - El Salvador: 19.0; 2.0; -1.8; -4.5; -5.2
  - Guatemala: 12.8; 2.2; -3.1; -8.1; -15.0
  - Honduras: 19.4; 1.9; -3.8; -6.6; -10.6
  - Nicaragua: 8.6; 0.8; -3.9; -11.5; -14.8
  - Panama: 161.8; 51.9; 22.8; 10.6; -3.0
  - EU-15: 92.2; 39.6; 7.9; -8.0; -20.4
  - Denmark: 164.5; 31.3; -19.6; -33.2; -44.2
  - Ireland: 52.5; 78.3; 12.9; -6.8; -21.9
  - Sweden: 114.1; 33.9; -6.5; -17.8; -28.9

### Simulated tax-financed increase in social spending (Table 9)
- Scenario design: increase social spending by 1 percent of GDP financed by equivalent tax increase; four permutations:
  1. Increase in all taxes proportional to current shares, proceeds distributed according to current incidence of social spending.
  2. Increase in all taxes proportional to current shares, proceeds distributed equally to everyone.
  3. Increase in VAT only, proceeds distributed according to current incidence.
  4. Increase in VAT only, proceeds distributed equally to everyone.
- Key reported changes in Gini (Quasi-Gini change = post-reform minus pre-reform, times 100) — selected results:
  - Simulation 1 (overall tax, current incidence):
    - Costa Rica: 3.3; 1.4; 0.6; -0.1; -0.9; -0.6
    - El Salvador: 4.4; 1.3; 0.4; -0.2; -0.5; -0.4
    - Guatemala: 3.7; 2.1; 1.1; 0.2; -1.1; -0.7
    - Nicaragua: 3.3; 2.1; 1.3; 0.1; -0.8; -0.6
    - Panama: 5.4; 2.7; 1.2; 0.4; -1.2; -0.8
  - Simulation 2 (overall tax, evenly channeled):
    - Costa Rica: 2.5; 1.1; 0.7; 0.0; -0.8; -0.5
    - El Salvador: 3.2; 0.9; 0.3; -0.1; -0.4; -0.3
    - Nicaragua: 6.1; 2.7; 1.2; 0.1; -1.1; -0.8
    - Panama: 5.5; 2.3; 1.2; 0.4; -1.1; -0.7
  - Simulation 3 (VAT only, current incidence):
    - Costa Rica: 3.3; 1.4; 0.5; -0.2; -0.8; -0.6
    - El Salvador: 3.9; 1.1; 0.3; -0.2; -0.4; -0.4
    - Nicaragua: 2.2; 1.4; 0.9; -0.2; -0.4; -0.4
    - Panama: 5.4; 2.5; 0.9; 0.1; -1.0; -0.7
  - Simulation 4 (VAT only, evenly channeled):
    - Costa Rica: 2.5; 1.1; 0.6; 0.0; -0.7; -0.5
    - El Salvador: 2.7; 0.7; 0.2; -0.1; -0.3; -0.2
    - Guatemala: 4.2; 1.7; 0.6; -0.1; -0.8; -0.6
    - Nicaragua: 5.1; 2.0; 0.7; -0.2; -0.7; -0.6
    - Panama: 5.4; 2.0; 0.8; 0.1; -0.9; -0.7
- Aggregate simulation findings:
  - A 1 percent of GDP increase in tax revenue devoted to social spending reduces the income Gini coefficient between 0.5 and 0.8 percentage points in Costa Rica, Guatemala, Nicaragua, and Panama, and between 0.2 and 0.4 percentage points in El Salvador and Honduras.
  - Improving targeting of social spending can considerably increase redistribution (example: Nicaragua’s impact on the poorest quintile would double if regressivity were improved to at least a flat distribution).
  - The redistributive impact is not much affected by whether financing comes from VAT or other taxes, even if VAT is regressive in isolation.

### Conclusions and policy recommendations
- Taxation has limited redistributive potential compared with social spending; the latter is the more powerful tool for redistribution in Central America.
- Policy implications:
  - Focus tax policy on raising revenue efficiently to expand the pool available for redistribution through social spending.
  - Consider broadening the tax base—even if that implies eliminating progressive exemptions—to increase resources for social spending and enhance overall progressivity.
  - Retain equity considerations in tax design: poor design (e.g., exemptions favoring richer segments) can harm equity, efficiency, effectiveness, and administration.
  - Increase the absolute level of public social spending in the region; current levels are generally low relative to GDP and total public spending.
  - Improve targeting of social spending:
    - Health and primary education are already strongly progressive and should be sustained.
    - Pensions and tertiary education are very regressive; access and coverage should be improved to enhance impact on the poor.
    - Expand well-targeted social assistance (e.g., conditional cash transfers) to reduce poverty, especially over the long run.
- Overall: combined taxation and well-targeted social spending can substantially improve incomes of the poor, even when the tax system alone is regressive.

*Source: _wp10112 - 2004. For Europe, 2001data.*

### Appendix 1: On What Basis Should the Tax Burden Be Measured?

### Appendix 1: On What Basis Should the Tax Burden Be Measured?

### Income vs. Consumption: conceptual trade-offs
- Progressivity and regressivity are defined relative to a welfare measure that proxies household capacity to pay taxes; the traditional measure is current income per household.
- Problems with current income as a welfare measure:
  - It is volatile and subject to temporary shocks; a survey over a particular period ignores the household’s position in its life cycle. Ideally capacity to pay should be measured relative to permanent or lifetime income.
  - Certain income types tend to be under-represented in surveys, particularly income from self-employment, professional services, and capital (interest, dividends).
  - Inheritances, transfers, and family remittances are often poorly captured in survey-based household income measures; this is a particular concern in Central America where family remittances are an important source of income and welfare, especially for the poor.
  - Current income does not consider the number and age of household members, which affect welfare and capacity to pay for a given income. (footnote in source)

- Rationale for using consumption as an alternative:
  - Consumption is less volatile than current income and can be a reasonable proxy for permanent income.
  - Consumption is less likely to be under-reported and typically reflects donations, remittances, and other transfers even when those are not fully captured in income data.
  - Because consumption tends to be more evenly distributed than income, incidence studies using consumption often find overall taxation, and consumption-based taxes in particular, to be more progressive than studies using current income (citing Fullerton and Rogers, 1993).

### Empirical evidence from Central America (as reported)
- Studies that use consumption as the welfare metric tend to find greater progressivity of taxation than income-based studies; this pattern is observed for:
  - El Salvador, Nicaragua, Panama (Appendix Table A.4)
  - Guatemala (Auguste and Artana, 2005)
  - Honduras (Gillingham, Newhouse, and Yakovlev, 2008)

### Limitations and practical constraints of using consumption
- Conceptual limitations:
  - Consumption may be a deficient measure of permanent income in the presence of bequest motives or precautionary savings; present savings cannot always be clearly interpreted as future consumption.
  - Empirically, richer households are found to permanently consume a lower share of their income than poorer households, even at later life-cycle stages.

- Data availability constraints:
  - Many household surveys do not measure consumption; for cross-country comparisons this limits the analyst’s ability to use consumption.
  - The use of income in this paper was forced by that constraint.

### Methodological references and sources used in the appendix
- This appendix is based mainly on Barreix, Roca, and Villela (2006) and Auguste and Artana (2005).
- Additional methodological and empirical references cited include Poterba (1989), Fullerton and Rogers (1993), Barthold (1993), Metcalf (1994), and a range of country studies and analytical papers listed in the appendix reference list.

*Source: IMF staff compilation from "Appendix 1: On What Basis Should the Tax Burden Be Measured?"*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp10112.pdf_