1. Research on redistributive effects of Croatian fiscal system Ivica Urban Institute of Public Finance

2. Distribution of fiscal burdens and benefits in Croatia Only partially investigated Distribution of PIT analyzed in several studies Distribution of VAT Influence of social benefits on income inequality

3. Distribution of fiscal burdens and benefits in Croatia No attempt to provide combined incidence calculation for several tax and benefit instruments

4. Income inequality in Croatia World Bank (2006), Nestić (2005), World Bank (2001) Disposable income inequality in Croatia is low in comparison with other transition countries What is the influence of government on distribution of income?

5. Theses for research Government redistributive policies are one of the prime determinants of disposable income inequality in Croatia The largest part of income redistribution process is achieved through the public pension system Fiscal instruments are highly efficient in reducing inequality (well-targeted and introducing low horizontal inequalities)

6. “Inequality” studies Kesselman & Cheung (2004): Tax Incidence, Progressivity, and Inequality in Canada Computable general equilibrium (CGE) Fiscal incidence (FINC) and Inequality (INEQ) studies: analyze single tax forms or groups of taxes and/or benefits

7. “Inequality” studies Assumptions (deficiencies, restrictions, problems with...) a) inclusion only of direct taxes and benefits in analysis; exclusion of indirect taxes and in- kind government benefits, b) income earners cannot shift the burden of personal taxes, c) existence of taxes and transfers does not affect market incomes, d) annual instead of lifetime perspective

8. “Inequality” studies: examples Kim & Lambert (2007): Redistributive Effect of US Taxes and Public Transfers, 1994-2004 Immervoll et al. (2005): Household incomes and redistribution in the European Union: quantifying the equalising properties of taxes and benefits

9. Methodology = Kakwani progressivity index of taxes V = Vertical effect R = Reranking effect G X = Gini coefficient of pre-fiscal income G N = Gini coefficient of post-fiscal income RE = Redistributive effect = Kakwani regressivity index of benefits Lambert (1986, 2001): Immervoll et al. (2005): D N = Concentration coefficient of post-fiscal income

10. Alternative approaches Methods: 1. Lambert (1986) – Decomposition of vertical effect 2. Immervoll et al. (2005) – “Exclusion” of one-by-one instrument from the reference base Reference income base: 1. Pre-TB income 2. Post-TB income

11. Equivalent incomes

12. Defining groups of households EM = employed or self-employed, UN = unemployed or inactive; EL = the elder

13. Distribution of income, taxes and benefits in 2005 (1)

14. Distribution of income, taxes and benefits in 2005 (2)

15. Entropy index decomposition

16. Distribution of income, taxes and benefits in 2005 (3)

17. Inequality measures

18. Decomposition of redistributive effect

19. Preliminary conclusions Croatian system of direct taxes, pensions and social benefits seems to be highly redistributive Public pensions are the most contributive instrument, followed by SSC and PIT Unfinished research, still lot to do

20. Fiscal system coverage PIT and SSC = 38% of total general government revenue Public pensions and cash benefits = 34% of total expenditures (2006) Huge part of fiscal system is still outside the coverage of this analysis

21. Further issues (1) How does Croatian fiscal system stand in comparison with EU countries in terms of achieving equity goals? Are the individual instruments and the system as a whole enough redistributive? Are there some instruments that are not equitable? Should taxes be more progressive? Can the overall redistributive effect be significantly altered through increase (decrease) of the PIT rate schedule progressivity?

22. Further issues (2) How can total welfare be increased keeping the amount of expenditures and taxes unchanged? What equity role have public pensions now, and how it should be in future? How can other fiscal instruments be included into current model to obtain more comprehensive picture of fiscal incidence?