## 1. Tax Variables ## Source details **Canonical URL:** [1. Tax Variables](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09136.pdf) ## Other formats - [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09136.pdf.md) - [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09136.pdf.json) --- ### I. Introduction - Tax incentives are common, especially in developing countries; empirical evidence is limited. - Existing empirical work categories: - Case studies (single-country, limited generalizability). - Calculations of effective marginal tax rates (reveal incentives at the margin, not outcomes). - Econometric studies on specific incentives in advanced economies (e.g., R&D tax credits, enterprise zones). - Paper objectives: - Test whether countries compete over tax incentives (tax holidays and investment allowances) as they do over CIT rates. - Assess whether these incentives affect FDI, gross private investment, and economic growth. - Key high-level findings: - Fiscal interdependence exists for CIT rates in the sample of developing countries; tax holidays also show interdependence. - Tax holidays affect FDI but there is no clear evidence of effects on total private investment or economic growth. - Countries appear to compete over instruments relevant to FDI, consistent with policy-makers taking firms’ perspectives into account. ### II. Data - Corporate tax data source: Price Waterhouse worldwide summaries of corporate taxes (published between 1985 and 2004). - Sample: initial unbalanced panel of 47 developing countries over 20 years, reduced by macro data availability. - Tax variables constructed: - CIT rate: manufacturing rate for the most profitable firms (if multiple rates exist). - Tax holidays: longest available holiday of the manufacturing or export sector in years. - Investment allowance/tax credit: most generous investment allowance of the manufacturing or export sector in percent of the investment; tax credit divided by tax rate for comparability. - Macroeconomic sources: IMF World Economic Outlook (WEO) and World Bank World Development Indicators (WDI). - Investment measures: - FDI (includes takeovers) and private gross fixed capital formation (“private investment”); both divided by GDP. - Control variables: inflation, general government consumption expenditure, GDP, population size, openness (Squalli & Wilson (2006) measure). - Descriptive statistics (preserve numeric values exactly): - CIT rate: Percent | Mean 34.69 | Std. Dev. 9.90 | Min 0.056 | Max 82.8 | Obs. 828 | Source PWC - Tax holiday: Years | Mean 3.5 | Std. Dev. 5.0 | Min 0.0 | Max 20 | Obs. 821 | Source PWC - Inv. allow./tax credit: Percent | Mean 22.6 | Std. Dev. 33.4 | Min 0.0 | Max 185.7 | Obs. 828 | Source PWC - FDI: Percent of GDP | Mean 2.9 | Std. Dev. 7.4 | Min -82.9 | Max 90.7 | Obs. 910 | Source WDI - Private Investment: Percent of GDP | Mean 14.3 | Std. Dev. 6.6 | Min 0.0 | Max 50.2 | Obs. 864 | Source WEO - Population: Millions | Mean 18.4 | Std. Dev. 30.2 | Min 0.1 | Max 184.3 | Obs. 940 | Source WEO - Openness (exports + imports / GDP): Percent of GDP | Mean 74.0 | Std. Dev. 41.0 | Min 12.1 | Max 250.0 | Obs. 925 | Source WEO - Openness (Sqalli & Wilson, 2006): Percent of GDP | Mean 20.0 | Std. Dev. 52.5 | Min 0.9 | Max 571.2 | Obs. 925 | Source WEO - GDP growth: Percent | Mean 3.0 | Std. Dev. 6.4 | Min -51.0 | Max 106.3 | Obs. 932 | Source WDI - GDP per capita: 1000s const. U.S. dollars | Mean 2.5 | Std. Dev. 2.8 | Min 0.1 | Max 16.5 | Obs. 933 | Source WDI - GDP: Bn. const. U.S. dollars | Mean 42.9 | Std. Dev. 112.3 | Min 0.1 | Max 717.3 | Obs. 933 | Source WDI - Gov. Consumption: Percent of GDP | Mean 14.6 | Std. Dev. 5.7 | Min 2.9 | Max 43.5 | Obs. 901 | Source WDI - Inflation: Percent | Mean 135.0 | Std. Dev. 1152.0 | Min -29.2 | Max 26762.0 | Obs. 932 | Source WDI - Time patterns (balanced panel of 29 countries): - Downward trend in statutory CIT rates. - Tax holidays have become shorter and investment allowances smaller; value of allowances to taxpayers fell even more because allowance × tax rate declined. ### III. Empirical Evidence on Tax Competition (Methodology and Fiscal Interactions) - Spatial lag model specification: - y_it = ρ W y_it + X_it β + η_i + ε_it (transformed into deviations from means in estimation). - Wy is the spatial lag (W weighting matrix); ρ is interaction coefficient. - Weighting: inverse distance weight matrix used. - Controls: GDP per capita, population, openness (Squalli & Wilson), general government consumption (lagged by one period). - Endogeneity concerns and estimation choice: - Spatially lagged variable endogenous → maximum likelihood estimator for spatial lag model used (Anselin 1988; Elhorst 2003). - Spatial error model also estimated; robust Lagrange Multiplier (LM) tests used to choose between lag vs. error dependence. - Balanced panel construction: - Main balanced subsets: 38 countries over 11 years (1988–98) and 37 countries over 12 years (1991–2002). Main results based on 38×11 set. - Time effects: time trend included (instead of year dummies) to avoid multicollinearity with weighted averages. - Key estimation results (preserve numeric magnitudes and reported standard errors): - Econometric method comparison (selected Wy (ρ) estimates and standard errors): - Wy (ρ) for CIT rate: OLS 0.395*** (0.111); MLSL 0.277*** (0.096); MLSE 0.619*** (0.135) - Wy (ρ) for Holiday: OLS 0.363*** (0.100); MLSL 0.346*** (0.080); MLSE 0.359*** (0.081) - Fiscal interactions without time trend (interpretation from coefficients): - Country reacts to a 10 percentage point drop in average CIT rate of others by lowering its own rate by 2.77 percentage points. - Country reacts to a decline of the average tax holiday of one year by reducing its own holiday by just over 4 months (0.36 years). - No evidence of spatial interaction on investment allowance. - Fiscal interactions with time trend: - Inclusion of time trend renders the interaction coefficients for CIT and tax holiday insignificant. - Time trend coefficients reported: CIT rate trend -0.589*** (0.131); Holiday trend -0.319*** (0.067); Inv. allow. trend -0.023*** (0.003). - Control-variable patterns: - As countries become richer they tend to use lower tax rates and more generous allowances (not fully consistent across specifications). - Larger populations associated with lower CIT rates and more generous incentives. - Government consumption associated with higher CIT rates. - Summary interpretation: - Evidence of interaction for tax rates and tax holidays but not for investment allowances. - Cannot conclusively distinguish between resource-flow (competition for mobile capital) and yardstick/spillover explanations from these fiscal-interaction results alone. ### IV. Empirical Evidence on the Effect on Investment and Growth - Dynamic panel model: - Inv_it = γ Inv_{i,t-1} + δ Tax_{i,t-1} + β X_{i,t-1} + μ_t + η_i + ε_it - Inv: FDI/GDP or private gross fixed capital formation/GDP; growth also considered. - Tax variables and controls lagged by one year. - Estimation approaches and selection: - OLS, within-groups (fixed effects), difference GMM (Arellano-Bond), system GMM (Blundell-Bond). - System GMM preferred; within-groups also reported. - Sample adjustments for specific regressions: - Bolivia and Venezuela dropped (extractive industries dominate); Liberia excluded; FDI regressions exclude Namibia; private investment regressions exclude Bahamas and Jamaica but include Namibia. - FDI results (preserve numeric magnitudes and standard errors where given): - Lagged FDI coefficient estimates: OLS 0.644*** (0.05); within-groups 0.353*** (0.06); difference GMM 0.220*** (0.07); system GMM 0.488*** (0.07). - System GMM instrument validity: Hansen J test p-value 1.000 reported. - Tax variable effects on FDI: - CIT rate: - OLS: -0.045*** (0.01) - Within groups: -0.033* (0.02) - Difference GMM: -0.060** (0.03) - System GMM: -0.045*** (0.02) - Interpretation: a 10 percentage point increase in the CIT rate lowers FDI by between 0.33 percentage points of GDP (within groups) and 0.45 percentage points of GDP (system GMM). - Tax holiday: - OLS: 0.069*** (0.02) - Within groups: 0.001 (0.02) - Difference GMM: 0.081** (0.04) - System GMM: 0.102*** (0.03) - Interpretation: adding ten years of tax holidays increases FDI by about 1 percent of GDP (system GMM); within-groups insignificance implies effect comes from cross-country variation rather than within-country changes. - Investment allowance: - Insignificant across estimators (example: system GMM coefficient -0.380 (0.40), not significant). - Other controls: mixed results; inflation small positive effect in some specifications; openness and GDP coefficients vary. - Private investment and growth results (preserve numeric magnitudes where reported): - Private investment persistence: - Within-groups lag coefficient 0.558***. - System GMM lag coefficient 0.634***. - Tax variables on private investment: - CIT rate: within-groups 0.056* (0.029) but system GMM 0.014 (0.032) — effect not robust. - Tax holiday: within-groups -0.054 (0.037); system GMM 0.083 (0.051) — not consistently significant. - Investment allowance: not significant. - Real economic growth: - Lagged growth significant: within-groups 0.208**; system GMM 0.244**. - CIT rate: system GMM -0.064* (0.032) (negative effect reported); tax holidays and investment allowances generally insignificant for growth. - Interpretation and potential explanations for observed patterns: - Tax rates and tax holidays affect FDI but not (robustly) total private investment or growth. - Possible reasons: - Instruments may influence takeover-related FDI (ownership changes) rather than greenfield investment (net capital stock). - Crowding out: increased FDI may crowd out domestically-financed investment, leaving total investment unchanged. - Limited spillovers or absorptive capacity: FDI may not generate growth unless financial/institutional absorptive capacity exists. ### V. Conclusion and Policy Implications - Main conclusions: - Strategic interaction exists for CIT rates and tax holidays in the sample of developing countries, but not for investment allowances. - Tax holidays and CIT rates are associated with increases in FDI; investment allowances are not effective in this sample. - The pattern that only instruments affecting FDI are used in strategic interaction supports (though does not prove) a resource-flow interpretation (competition for footloose FDI) rather than pure tax mimicking. - No robust evidence that tax incentives increase total private investment or economic growth. - Policy implications highlighted: - Economists’ advice to favor investment allowances or accelerated depreciation over tax holidays may be resisted by governments because tax holidays better attract rent-earning, highly profitable FDI. - Alternative policy recommendation: reduced tax rates rather than tax holidays, since reduced rates would be valuable for highly-profitable investment. - Suggestions for future research: - Extend analysis to Asia. - Distinguish by sectors and by FDI source country, given availability of disaggregated data. ### Country sample and coverage (selected listing and coverage note) - Botswana: all - Argentina: All - Antigua and Barbuda: all - Cameroon: 88-02 - Bolivia: All - Bahamas: all - Congo, Rep.: 88-00 - Brazil: All - Barbados: all - Gabon: 88-98, 01-02 - Chile: All - Dominican Republic: all - Ghana: 91-04 - Colombia: All - Jamaica: all - Ivory coast: all - Costa Rica: All - St. Lucia: all - Kenya: all - Ecuador: All - Trinidad and Tobago: all - Liberia: 85-91 - El Salvador: 85-98 - Malawi: 85-02 - Guatemala: All - Morocco: all - Guyana: 88-04 - Mauritius: all - Honduras: 85-98 - Mozambique: 99-04 - Mexico: All - Namibia: 91-04 - Nicaragua: 85-86,92-97,01-02 - Nigeria: 84-02 - Panama: All - Senegal: 85-02 - Paraguay: All - South Africa: All - Peru: All - Swaziland: 91-96,01-04 - Uruguay: All - Tanzania: 88-04 - Venezuela: All - Uganda: 91-04 - Dem. Rep. of Congo: 85-96, 01-04 - Zambia: all - Zimbabwe: all - Note: 1/ “All” indicates 1985–2004. - Data source note: Authors’ calculation based mostly on data in Price Waterhouse Coopers (various years). *Source: _wp09136 - 1. Tax Variables (PDF).* ### 1. Tax Variables ....................................................................................................... ### 1. Tax Variables ### Major sections - 1. Tax Variables .........................................................................................................................5 - 2. Descriptive Statistics..............................................................................................................6 - 3. The Choice Between Estimation Methods...........................................................................12 - 4. Fiscal Interactions for Different Tax Instruments................................................................13 - 5. The Choice of Estimation Method .......................................................................................18 - 6. The Effects of Tax Instruments on Investment and Growth................................................20 ### Figures - 1. Average Tax Rates and Incentives Over the Years ...............................................................7 ### Appendixes - 1. Countries and Periods Covered in the Corporate Tax Dataset. ...........................................22 *Source: _wp09136 - 1. Tax Variables (PDF).* ### References.............................................................................................................. ### _wp09136 - References.............................................................................................................. ### I. Introduction - Tax incentives are common, especially in developing countries; empirical evidence is limited. - Existing empirical work categories: - Case studies (single-country, limited generalizability). - Calculations of effective marginal tax rates (reveal incentives at the margin, not outcomes). - Econometric studies on specific incentives in advanced economies (e.g., R&D tax credits, enterprise zones). - Paper objectives: - Test whether countries compete over tax incentives (tax holidays and investment allowances) as they do over CIT rates. - Assess whether these incentives affect FDI, gross private investment, and economic growth. - Key high-level findings: - Fiscal interdependence exists for CIT rates in the sample of developing countries; tax holidays also show interdependence. - Tax holidays affect FDI but there is no clear evidence of effects on total private investment or economic growth. - Countries appear to compete over instruments relevant to FDI, consistent with policy-makers taking firms’ perspectives into account. ### II. Data - Corporate tax data source: Price Waterhouse worldwide summaries of corporate taxes (published between 1985 and 2004). - Sample: initial unbalanced panel of 47 developing countries over 20 years, reduced by macro data availability. - Tax variables constructed (Table 1): - CIT rate: manufacturing rate for the most profitable firms (if multiple rates exist). - Tax holidays: longest available holiday of the manufacturing or export sector in years. - Investment allowance/tax credit: most generous investment allowance of the manufacturing or export sector in percent of the investment; tax credit divided by tax rate for comparability. - Macroeconomic sources: IMF World Economic Outlook (WEO) and World Bank World Development Indicators (WDI). - Investment measures: - FDI (includes takeovers) and private gross fixed capital formation (“private investment”); both divided by GDP. - Control variables: inflation, general government consumption expenditure, GDP, population size, openness (Squalli & Wilson (2006) measure). - Descriptive statistics (Table 2) — preserve numeric values exactly: - CIT rate: Percent | Mean 34.69 | Std. Dev. 9.90 | Min 0.056 | Max 82.8 | Obs. 828 | Source PWC - Tax holiday: Years | Mean 3.5 | Std. Dev. 5.0 | Min 0.0 | Max 20 | Obs. 821 | Source PWC - Inv. allow./tax credit: Percent | Mean 22.6 | Std. Dev. 33.4 | Min 0.0 | Max 185.7 | Obs. 828 | Source PWC - FDI: Percent of GDP | Mean 2.9 | Std. Dev. 7.4 | Min -82.9 | Max 90.7 | Obs. 910 | Source WDI - Private Investment: Percent of GDP | Mean 14.3 | Std. Dev. 6.6 | Min 0.0 | Max 50.2 | Obs. 864 | Source WEO - Population: Millions | Mean 18.4 | Std. Dev. 30.2 | Min 0.1 | Max 184.3 | Obs. 940 | Source WEO - Openness (exports + imports / GDP): Percent of GDP | Mean 74.0 | Std. Dev. 41.0 | Min 12.1 | Max 250.0 | Obs. 925 | Source WEO - Openness (Sqalli & Wilson, 2006): Percent of GDP | Mean 20.0 | Std. Dev. 52.5 | Min 0.9 | Max 571.2 | Obs. 925 | Source WEO - GDP growth: Percent | Mean 3.0 | Std. Dev. 6.4 | Min -51.0 | Max 106.3 | Obs. 932 | Source WDI - GDP per capita: 1000s const. U.S. dollars | Mean 2.5 | Std. Dev. 2.8 | Min 0.1 | Max 16.5 | Obs. 933 | Source WDI - GDP: Bn. const. U.S. dollars | Mean 42.9 | Std. Dev. 112.3 | Min 0.1 | Max 717.3 | Obs. 933 | Source WDI - Gov. Consumption: Percent of GDP | Mean 14.6 | Std. Dev. 5.7 | Min 2.9 | Max 43.5 | Obs. 901 | Source WDI - Inflation: Percent | Mean 135.0 | Std. Dev. 1152.0 | Min -29.2 | Max 26762.0 | Obs. 932 | Source WDI - Time patterns (Figure 1, balanced panel of 29 countries): - Downward trend in statutory CIT rates. - Tax holidays have become shorter and investment allowances smaller; value of allowances to taxpayers fell even more because allowance × tax rate declined. ### III. Empirical Evidence on Tax Competition H3: Methodology - Spatial lag model specification (equation (1)): - y_it = ρ W y_it + X_it β + η_i + ε_it (transformed into deviations from means in estimation). - Wy is the spatial lag (W weighting matrix); ρ is interaction coefficient. - Weighting: inverse distance weight matrix used (competition likely beyond first-order neighbors; incomplete dataset would leave few/neighbors otherwise). - Controls: GDP per capita, population, openness (Squalli & Wilson), general government consumption (lagged by one period). - Endogeneity concerns: - Spatially lagged variable endogenous → maximum likelihood estimator for spatial lag model (Anselin 1988; Elhorst 2003) used. - Also estimate spatial error model and use robust Lagrange Multiplier (LM) test to choose between lag vs. error dependence. - Balanced panel construction: - Tax-variable averages computed on a fully balanced panel to avoid composition changes; main balanced subsets: 38 countries over 11 years (1988–98) and 37 countries over 12 years (1991–2002). Main results based on 38×11 set. - Time effects: time trend included (instead of year dummies) to avoid multicollinearity problems with weighted averages. H3: Results - Econometric method choice (Table 3) — key points: - OLS shows strong fiscal interdependence but upward biased. - Maximum likelihood spatial lag (MLSL) estimates smaller but significant interaction coefficients. - Robust LM tests favor the spatial lag model for CIT rate; for tax holidays both lag and error tests admitted dependence but larger statistic indicates spatial lag. - Example coefficients (Table 3 highlights): - Wy (ρ) for CIT rate: OLS 0.395*** (0.111); MLSL 0.277*** (0.096); MLSE 0.619*** (0.135) - Wy (ρ) for Holiday: OLS 0.363*** (0.100); MLSL 0.346*** (0.080); MLSE 0.359*** (0.081) - Fiscal interactions (Table 4) — with and without time trend: - Without time trend (columns 1,3,5): - Country reacts to a 10 percentage point drop in average CIT rate of others by lowering its own rate by 2.77 percentage points. - Country reacts to a decline of the average tax holiday of one year by reducing its own holiday by just over 4 months (0.36 years). - No evidence of spatial interaction on investment allowance. - With time trend (columns 2,4,6): - Inclusion of time trend renders the interaction coefficients for CIT and tax holiday insignificant — identification between common declining trend and interaction difficult. - Time trend coefficients: CIT rate trend -0.589*** (0.131), Holiday trend -0.319*** (0.067), Inv. allow. trend -0.023*** (0.003). - Control variable patterns: - As countries become richer they tend to use lower tax rates and more generous allowances (not fully consistent across specifications). - Larger populations associated with lower CIT rates and more generous incentives. - Government consumption associated with higher CIT rates (CIT as a revenue source). - Summary interpretation: - Evidence of interaction for tax rates and tax holidays but not for investment allowances. - Cannot conclusively distinguish between resource-flow (competition for mobile capital) and yardstick/spillover explanations; subsequent investment analysis used to shed light. ### IV. Empirical Evidence on the Effect on Investment and Growth H3: Methodology - Dynamic panel model (equation (5)): - Inv_it = γ Inv_{i,t-1} + δ Tax_{i,t-1} + β X_{i,t-1} + μ_t + η_i + ε_it - Inv: FDI/GDP or private gross fixed capital formation/GDP; growth also considered. - Tax variables lagged by one year; controls lagged by one year. - Estimation approaches: - OLS, within-groups (fixed effects), difference GMM (Arellano-Bond), system GMM (Blundell-Bond). - System GMM preferred (performed best on lagged dependent variable coefficient and instrument validity); within-groups also reported given panel dimensions. - Data sample adjustments: - Bolivia and Venezuela dropped (extractive industries dominate); Liberia excluded for lack of macro data; regressions on FDI exclude Namibia; private investment regressions exclude Bahamas and Jamaica but include Namibia. H3: Results — FDI (Table 5) - Choice of estimator: - Lagged FDI coefficient estimates: OLS 0.644*** (0.05); within-groups 0.353*** (0.06); difference GMM 0.220*** (0.07); system GMM 0.488*** (0.07). - System GMM chosen for main inference; Sargan/Hansen tests do not reject instrument sets (Hansen J test p-value 1.000 reported for system GMM). - Tax variable effects on FDI (preserve numeric magnitudes): - CIT rate: - OLS: -0.045*** (0.01) - Within groups: -0.033* (0.02) - Difference GMM: -0.060** (0.03) - System GMM: -0.045*** (0.02) - Interpretation: a 10 percentage point increase in the CIT rate lowers FDI by between 0.33 percentage points of GDP (within groups) and 0.45 percentage points of GDP (system GMM). - Tax holiday: - OLS: 0.069*** (0.02) - Within groups: 0.001 (0.02) - Difference GMM: 0.081** (0.04) - System GMM: 0.102*** (0.03) - System GMM: adding ten years of tax holidays increases FDI by about 1 percent of GDP. - Within-groups insignificance implies effect comes from cross-country variation rather than within-country changes. - Investment allowance: - Insignificant across estimators (e.g., system GMM coefficient -0.380 (0.40), not significant). - Other controls: mixed results; inflation small positive effect in some specifications; openness and GDP coefficients vary. H3: Results — Private Investment and Growth (Table 6) - Private investment: - Lagged private investment significant and persistent (within-groups 0.558***; system GMM 0.634***). - Tax variables: - CIT rate: within-groups 0.056* (0.029) but system GMM 0.014 (0.032) — effect not robust; system GMM indicates no significant impact of CIT on total private investment. - Tax holiday: within-groups -0.054 (0.037); system GMM 0.083 (0.051) — not consistently significant. - Investment allowance: not significant. - Real economic growth: - Lagged growth significant (within-groups 0.208**; system GMM 0.244**). - CIT rate: negative effect on growth in system GMM: -0.064* (0.032) (also negative but not robust in some specs). - Tax holidays and investment allowances generally insignificant for growth. - Interpretation and potential explanations for patterns: - Tax rates and tax holidays affect FDI but not (robustly) total private investment or growth. - Possible reasons: - Tax instruments influence takeover-related FDI (ownership changes) rather than greenfield investment (net capital stock). - Crowding out: increased FDI may crowd out domestically-financed investment, leaving total investment unchanged. - Lack of spillovers/absorptive capacity: FDI may not generate growth unless financial/institutional absorptive capacity exists. ### V. Conclusion - Main conclusions: - Strategic interaction exists for CIT rates and tax holidays in the sample of developing countries, but not for investment allowances. - Tax holidays and CIT rates are associated with increases in FDI; investment allowances are not effective in this sample. - The pattern that only instruments affecting FDI are used in strategic interaction supports (though does not prove) a resource-flow interpretation (competition for footloose FDI) rather than pure tax mimicking. - No robust evidence that tax incentives increase total private investment or economic growth; possible implications include crowding out or effects concentrated in ownership changes rather than additional capital formation. - Policy implications highlighted: - Economists’ advice to favor investment allowances or accelerated depreciation over tax holidays may be resisted by governments because tax holidays better attract rent-earning, highly profitable FDI. - Alternative policy recommendations could include reduced tax rates rather than tax holidays, since reduced rates would be valuable for highly-profitable investment. - Suggestions for future research: - Extend analysis to Asia. - Distinguish by sectors and by FDI source country, given availability of disaggregated data. *Source: Authors’ calculations and analysis as presented in _wp09136 - References..............................................................................................................* ### 1. Botswana all Argentina All Antigua and ### _wp09136 - 1. Botswana all Argentina All Antigua and ### Country sample and coverage - Botswana: all - Argentina: All - Antigua and Barbuda: all - Cameroon: 88-02 - Bolivia: All - Bahamas: all - Congo, Rep.: 88-00 - Brazil: All - Barbados: all - Gabon: 88-98, 01-02 - Chile: All - Dominican Republic: all - Ghana: 91-04 - Colombia: All - Jamaica: all - Ivory coast: all - Costa Rica: All - St. Lucia: all - Kenya: all - Ecuador: All - Trinidad and Tobago: all - Liberia: 85-91 - El Salvador: 85-98 - Malawi: 85-02 - Guatemala: All - Morocco: all - Guyana: 88-04 - Mauritius: all - Honduras: 85-98 - Mozambique: 99-04 - Mexico: All - Namibia: 91-04 - Nicaragua: 85-86,92-97,01-02 - Nigeria: 84-02 - Panama: All - Senegal: 85-02 - Paraguay: All - South Africa: All - Peru: All - Swaziland: 91-96,01-04 - Uruguay: All - Tanzania: 88-04 - Venezuela: All - Uganda: 91-04 - Dem. Rep. of Congo: 85-96, 01-04 - Zambia: all - Zimbabwe: all - Note: 1/ “All” indicates 1985–2004. ### Data sources and calculation note - Source: Authors’ calculation based mostly on data in Price Waterhouse Coopers (various years). ### References cited (selected entries from the source list) - Altshuler, R. and T.J. Goodspeed, 2002, “Follow the Leader? Evidence on European and U.S. Tax Competition,” Rutgers University Department of Economics Working Papers 200226. - Anselin, L., 1988, “Spatial Econometrics: Methods and Models,” Hingham (Massachusetts: Kluwer Academic Publishers). - Anselin, L., A.K. Bera, R. Florax, and M.J. Yoon, 1996, “Simple Diagnostic Tests for Spatial Dependence,” Regional Science and Urban Economics, 26, pp. 77–104. - Arellano, M. and S. Bond, 1991, “Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application to Employment Equations,” Review of Economic Studies, 58, pp. 277–297. - Baldwin, R. and P. Krugman, 2004, “Agglomeration, Integration, and Tax Harmonization,” European Economic Review, 48, pp. 1–23. - Baltagi, B.H., P. Egger, and M. Pfaffermayr, 2007, “Estimating Models of Complex FDI: Are There Third-Country Effects?” Journal of Econometrics, 140, pp. 260–81. - Besley, T. and A. Case, 1995, “Incumbent Behavior: Vote-Seeking, Tax-Setting, and Yardstick Competition,” American Economic Review, 85, pp. 25–45. - Bird, R.M., 2000, “Tax Incentives for Investment in Developing Countries,” in Fiscal Reform and Structural Change in Developing Countries ed. by G. Perry, J. Whalley, and G. McMahon, Vol. 1, McMillan Press, pp. 201–21. - Bloom, N., R. Griffith, and J. Van Reenen, 2002, “Do R&D Tax Credits Work? Evidence from a Panel of Countries, 1979–1997,” Journal of Public Economics, Vol. 85, pp. 1–31. - Blundell, R. and S. Bond, 1998, “Initial Conditions and Moment Restrictions in Dynamic Panel Data Models,” Journal of Econometrics, 87, pp. 11–143. - De Mooij, R.A. and S. Ederveen, 2003, “Taxation and Foreign Direct Investment: A Synthesis of Empirical Research,” International Tax and Public Finance, Vol. 10(6), pp. 673–93. - Devereux, M. P., R. Griffith, and A. Klemm, 2002, “Corporate Income Tax Reforms and International Tax Competition,” Economic Policy, Vol. 17(35), pp. 451–495. - Devereux, M. P., B. Lockwood, and M. Redoano, 2008, “Do Countries Compete Over Corporate Tax Rates?” Journal of Public Economics, Vol. 92, pp. 1210–35. - Klemm, A., 2009, “Causes, Benefits and Risks of Tax Incentives,” IMF Working Paper 09/21 (Washington: International Monetary Fund). - Price Waterhouse/PriceWaterhouseCoopers, various years, “World Corporate Tax Guide.” - Zee, H.H., J.G. Stotsky, and E. Ley, 2002, “Tax Incentives for Business Investment: A Primer for Policy Makers in Developing Countries,” World Development, Vol. 30(9), pp. 1497–1516. *Source: Authors’ calculation based mostly on data in Price Waterhouse Coopers (various years). 1/ “All” indicates 1985–2004.* --- _Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp09136.pdf_