This thesis aims to answer two questions utilizing Norwegian data for 2007; 1) what is the distribution of taxes and public spending across different income deciles and household types? and 2) taking this into account, how does it affect income inequality and poverty?
The first step in answering the two questions is to estimate the distribution of taxes. While all direct taxes are obtained immideatly from data, indirect taxes requires a different approach. Data from the Consumer Survey's of Statistics Norway is used together with effective tax shares of consumer prices to find indirect taxes paid. When looking at income deciles, consumption expenditures are not known, so these are estimated in an Almost Ideal Demand System. The last tax distributed is the payroll tax, under the assumption of inelastic labor supply. Corporate taxes are left out of the analysis because of no clear consesus regarding the incidence assumptions. Summing the different taxes together, I find the distribution of taxes as share of income for all groups, where the total average tax rate is 46,3%.
Moving on to public expenditures, these are distributed in two steps. First, direct cash transfers are obtained and distributed directly from available data. The value of public services is distributed according to an allocation rule derived from a theoretical framework. All municipal- and county-provided services are covered. In addition to these, I also impute the value of public hospitals. The value of public services is assumed to be equal the cost of providing them. Since the allocation of public spending is likely to reflect the allocation of needs in the population, these values are needs-adjusted according to estimated equivalence scales drawn from the same model which was used to derive the allocation rule. Finally, the fiscal incidence - the incidence of all (covered) taxes and all (covered ) public expenditures is estimated for all household types and income deciles.
I find that there are large variations in the fiscal incidence across groups. Typically “weak” (retiree-abundant groups and lone parents) and poor households recieve large amounts of public expenditures relative to what they pay in taxes. This is a reflection of the redistributional properties of the fiscal budget. There are also relatively large differences in the fiscal incidence for certain groups before and after needs-adjusting. For households who are not a typical risk-group in terms of poverty but still recieves a large amount of public spending, such as couples with small children, the fiscal incidence changes from being a net recipient of public services to being a net financieer when needs-adjusting.
Knowing the incidence of taxes and public spending, I move on to answering the second question; does this change the degree of economic inequality? For instance, since public expenditures cater to many poor groups, what happens to economic inequality when recieved public goods are added as an income component?
Comparing the Gini-indicies for different income measures I find that the traditional approach (estimating a Gini-coefficient for the distribution of cash income net of direct taxes) produces a higher measured inequality than when accounting for the provision of public services and additional taxes. Accounting for indirect taxes, the payroll tax and in-kind public services, the Gini-coefficient is reduced from approximately 0.237 to 0.182. I also find that the poverty incidence is reduced by 11 - 13% when taking the fiscal incidence into account.