Fiscal incidence remains an important and intriguing topic. Despite the fact that fiscal incidence is often at the center of public debate, analytical work on fiscal incidence has been sparse since the seminal contribution of Aaron and McGuire (1970). This is especially so for the benefit side of public budgets. The purpose of this paper is to contribute to the methodology of fiscal incidence by studying analytically the definition and measurement of the incidence of an expenditure program when welfare is used as a metric, when benefits may be shifted, and when price changes are important. The paper first addresses the conceptual steps required for analytical work on incidence: the choice of a counterfactual with which to define program impacts; the assumption about the nature of financing of the program under examination; the choice of a metric with which incidence is to be measured; the choice of benchmark relative to which changes in benefits are to be compared for the purpose of generating incidence across groups; and the definition of shifting and the role of behavioral assumptions therein. We consider each of these steps in turn, with special attention to comparison of analyses that alternatively use welfare or budgetary amounts as a metric.
To make the discussion concrete, we then carry out these conceptual steps in the context of an analysis of a particular benefit program: a tax-financed (price) subsidy for home care for elderly parents, when their children are altruistic. This is a policy that is likely to grow in importance because of the aging populations of western countries.
In the course of the application, we consider the conditions under which welfare and budgetary analyses converge. Using simulation, we also consider the importance of behavioral assumptions about family structure for the analysis of the shifting of benefits, in a manner analogous to the investigation of the role of behavioral assumptions in the analysis of tax shifting.