Rossana Merola ILO, Research Department

Raising tax revenues without harming equity and employment Rossana Merola ILO, Research Department G-24 Special Workshop on Growth and Reducing Inequality Geneva, 5 th - 6 th September 2017 The views expressed in this paper are those of the authors and do not necessarily represent those of the ILO R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 1 / 23

Introduction - Motivations Debt-to-GDP ratios rose in aftermath of crisis AUS AUT BEL CAN 40 60 80 60 80 100 100 120 140 80 100 120 2005 2010 2015 2005 2010 2015 2005 2010 2015 2005 2010 2015 DEU DNK ESP FIN 60 70 80 90 30 40 50 60 50 100 150 40 60 80 2005 2010 2015 2005 2010 2015 2005 2010 2015 2005 2010 2015 FRA GBR IRL ITA 80 100 120 50 100 150 50 100 150 120 140 160 2005 2010 2015 2005 2010 2015 2005 2010 2015 2005 2010 2015 JPN PRT SWE USA 150 200 250 50 100 150 50 55 60 65 80 100 120 2005 2010 2015 2005 2010 2015 2005 2010 2015 2005 2010 2015 Source: 2016 OECD Economic Outlook R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 2 / 23

Introduction - Motivations Governments started consolidating deficit-to-gdp AUS AUT BEL CAN -5 0 5-6 -4-2 0-6 -4-2 0-4 -2 0 2 2005 2010 2015 2005 2010 2015 2005 2010 2015 2005 2010 2015 DEU DNK ESP FIN -4-2 0-5 0 5-10 -5 0-5 0 5 2005 2010 2015 2005 2010 2015 2005 2010 2015 2005 2010 2015 FRA GBR IRL ITA -8-6 -4-2 -10-5 0-40 -20 0-6 -4-2 2005 2010 2015 2005 2010 2015 2005 2010 2015 2005 2010 2015 JPN PRT SWE USA -10-5 0-10 -5 0-2 0 2 4-15 -10-5 2005 2010 2015 2005 2010 2015 2005 2010 2015 2005 2010 2015 Source: 2016 OECD Economic Outlook R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 3 / 23

% TOTAL REVENUES Introduction - Motivations Taxation: advanced VS developing countries 40 35 30 25 20 15 10 5 0 TAX COMPOSITION, 2013 Taxes on goods and services High income countries Low & middle income countries Taxes on income, profits and capital gains Taxes on international trade Source: World Bank, World Development Indicators R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 4 / 23

Introduction - Motivations Consolidation may exacerbate income disparities in AEs AUS AUT BEL CAN 40 45 50 55 35 40 45 50 40 45 50 40 45 50 1980 1988 1996 2004 2012 1980 1988 1996 2004 2012 1980 1988 1996 2004 2012 1980 1988 1996 2004 2012 DEU DNK ESP FIN 40 45 50 40 45 50 55 35 40 45 50 40 45 50 1980 1988 1996 2004 2012 1980 1988 1996 2004 2012 1980 1988 1996 2004 2012 1980 1988 1996 2004 2012 FRA GBR IRL ITA 44 46 48 50 40 45 50 55 45 50 55 40 45 50 1980 1988 1996 2004 2012 1980 1988 1996 2004 2012 1980 1988 1996 2004 2012 1980 1988 1996 2004 2012 JPN PRT SWE USA 30 40 50 45 50 55 40 45 50 40 45 50 1980 1988 1996 2004 2012 1980 1988 1996 2004 2012 1980 1988 1996 2004 2012 1980 1988 1996 2004 2012 Source: Market Gini index-standardized Wealth and Inequality Database R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 5 / 23

Introduction - Motivations Inequality is rising also in some emerging countries CHN COL 30 40 50 60 45 50 55 1980 1988 1996 2004 2012 1980 1988 1996 2004 2012 MEX ZAF 44 46 48 50 60 65 70 1980 1988 1996 2004 2012 1980 1988 1996 2004 2012 Source: Standardized Wealth and Inequality Database R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 6 / 23

Introduction - Motivations Effects of consolidation on inequality 1 Direct effect through income redistribution Lower govt. transfers penalises poorer agents Inequality Higher top marginal income tax penalises richer agent Inequality 2 Indirect effect through distortions of labor supply decisions Substitution effect: higher income taxes discourage labour participation Income effect: higher taxes reduce purchasing power and encourage labour participation 3 General equilibrium effects through response of the economy during recessions Tendency of employers to hoard high-skilled workers (Bastagli et al., 2012) Inequality Capital owners may lose income in countries with rigid labour markets Inequality R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 7 / 23

Related literature and our contribution Research questions In this context, our analysis is two-fold: 1 What effects do tax-based consolidations have on employment, growth and income inequality? 2 Does it matter which tax instruments are being used? R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 8 / 23

Related literature and our contribution Related literature Effect of overall fiscal consolidations on income inequality Ball et al. (2013) & Woo et al. (2013) & Agnello and Sousa (2014) spending-based consolidations increase inequality tax-based consolidations increase/do not affect/decrease inequality do not go beyond tax vs spending-based distinction do not explore transmission channels Effect of budget-neutral changes in composition of taxes on: 1 Inequality: Martinez-Vasquez et al. (2012) do not address endogeneity between tax revenues and GDP pool together high, middle and low-income countries 2 GDP and inequality: Muinello-Gallo and Roca-Sagales (2013) do not address endogeneity between tax revenues and GDP R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 9 / 23

Related literature and our contribution Our contribution We focus only on exogenous tax shocks We start form the IMF narrative consolidation dataset compiled by Devries et al. (2011) and updated by Alesina et al. (2015) Identification of consolidation years with the only objective of reducing the budget in 16 OECD countries during 1978-2013. No output stabilization, nor labor conditions or inequality considerations. We analyse tax composition effects The IMF dataset does not provide info on tax composition we match IMF consolidation years with OECD Revenue Statistics Explore composition effects of tax-based consolidations on income inequality, while accounting for potential general equilibrium dynamics through economic activity and the labour market explain transmission channels PVAR R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 11 / 23

Related literature and our contribution Rest of the presentation 1 Overal effect of tax-based consolidation: 1% GDP change in total total tax revenues 2 Composition effects: Effect of 1% GDP change in direct tax revenues (not significant) Personal income tax Corporate tax and SSC (smaller effect) Effect of 1% GDP change in indirect tax revenues General taxes: VAT and sales Taxes on specific goods and services R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 12 / 23

Results: Overall effects of tax-based consolidation Overall, there is a trade-off between equity and growth Figure: IRF to a 1% of GDP total tax shock robustness Percent response of gross domestic product 0.5 0.0-0.5-1.0-1.5-2.0-2.5 Percentage point response of unemployment rate.8.6.4.2.0 -.2 -.4.2.1.0 -.1 -.2 -.3 -.4 -.5 -.6 -.7.3.2.1.0 -.1 -.2 -.3 -.4 Unit response of net Gini index Pecentage point response of participation rate Note: Blue line represents non-cumulated responses to a 1% of GDP change in total tax revenues, red lines represent 90% confidence intervals. R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 13 / 23

Results: Composition effects of tax-based consolidation Among direct taxes, PIT have the most significant effects Figure: IRF to a 1% of GDP personal income tax (PIT) shock Percent response of Gross Domestic Output 1 0-1 -2-3 Percentage point response of net Gini index 0.4 0.0-0.4-0.8-1.2-4 -1.6 Percentage point response of Unemployment rate 2.0 1.6 1.2 0.8 0.4 0.0-0.4-0.8 Percentage point response of Participation rate.6.4.2.0 -.2 -.4 -.6 -.8 R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 14 / 23

Results: Composition effects of tax-based consolidation Indirect taxes are more successful in reducing inequality Figure: IRF to a 1% of GDP indirect tax shock robustness 1 0-1 -2-3 -4-5 -6-7 -8 4 3 2 1 0 Percent response of Gross Domestic Output Percentage point response of Unemployment rate Percentage point response of net Gini index 0.4 0.0-0.4-0.8-1.2-1.6-2.0-2.4 Percentage point response of Participation rate 1.2 0.8 0.4 0.0-0.4-0.8-1 -1.2 Why? Short-run increase in labour force participation R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 15 / 23

Results: Composition effects of tax-based consolidation The labour supply channel is stronger for women Figure: IRF to a 1% of GDP indirect tax shock by gender and age GDP Gini coefficient 2 0.4 0 0.0-2 -0.4-4 -0.8-6 -1.2-8 -1.6-10 -2.0 Male employment rate (35-44) Male participation rate (35-44) 1 1 0 0-1 -1-2 -2-3 -4-3 -5-4 Female employment rate (35-44) Female participation rate (35-44) 1.2 2.5 0.8 2.0 1.5 0.4 1.0 0.0 0.5-0.4 0.0-0.5-0.8-1.0-1.2-1.5 R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 16 / 23

Results: Composition effects of tax-based consolidation GT are the most effective sub-component Figure: IRF to a 1% of GDP general tax (GT) shock Percent response of Gross Domestic Output 1 0-1 -2-3 -4-5 -6-7 Percentage point response of Unemployment rate 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0-0.5 Percentage point response of net Gini index 0.4 0.0-0.4-0.8-1.2-1.6-2.0 Percentage point response of Participation rate 1.6 1.2 0.8 0.4 0.0-0.4-0.8 Income effect dominates substitution effect participation inequality R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 17 / 23

Conclusions Conclusions for the 16 OECD countries Tax-based consolidations reduce inequality, but at cost of lower output and employment (although not significant) Trade-off Design of consolidation matters: Among direct taxes, only personal taxes reduce inequality Indirect taxes (and especially GT) create incentives for (poorer) agents to increase labour supply low income inequality Indirect taxes brings middle-aged women in the labor force less gender inequality R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 18 / 23

Conclusions Are our conclusions valid also for developing countries? Limitations: Direct taxes (PIT) Lobbying by economic groups limits the effectiveness of income taxes in reducing inequality High level of informality contributes to the problem of insufficient income tax revenues Indirect taxes (GT) The labour supply channel may still work, but it could encourage participation in the informal labour market rather than in the formal one R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 19 / 23

Conclusions Thank you very much! R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 20 / 23

Appendix Methodology back Panel Vector Auto-regressive (PVAR): A 0 Z i,t = A 1 Z i,t 1 + A 2 Z i,t 2 + α i + δ t + τ it + ɛ i,t (1) where Z i,t contains (in this order): (1) tax shock (2) real per capita GDP (in logs) (3) GINI coefficient (in % points) (4) unemployment rate (in % points) (5) labor force participation rate (in % points) We include: country fixed effects (α i ), country-specific linear trends (τ it ), year fixed effects (δ t ), and two lags. Identification through Cholesky decomposition ordering below shock does not matter R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 21 / 23

Appendix Robustness checks on baseline specification back 1 Type of consolidation (spending-based, unanticipated tax-based, tax-based with positive realization) 2 Alternative tax data (IMF narrative data, OECD cylically adjusted revenues) 3 Alternative inequality measure (WID Top income shares) 4 Reverse causality issues 5 Local projections method 6 Deterministic components (common linear trend, first differences) 7 Sample selection (no financial crisis, EU-only, no banking crises, no outliers) 8 Country sample stability (dropping one country at a time) 9 Control variables R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 22 / 23

Appendix Robustness check on tax composition back Possible criticism is inclusion of 1 shock at a time Address it by estimating two new specifications: with direct (ordered 1st) and indirect (ordered 2nd) tax shocks with indirect (ordered 1st) and direct (ordered 2nd) tax shocks and then for each VAR estimate IRFs to both tax shocks Country sample stability Top income shares R. Merola (ILO, Research Department) G24 Special Workshop 6 th September 2017 23 / 23