The Effect of State Tax Preferences on the Living Arrangements and Welfare of the Elderly

The United States is in the midst of profound demographic changes: the proportion of the elderly population (65+) in the US is projected to increase from a current 13% to 20% by 2030. Since the source of elderly income comes largely from social security and retirement accounts, state's income tax policies towards the elderly can have significant impact over the welfare of the aging population. While every state has a personal income tax system that offers some tax preferences to the elderly population, the policy forms and generosity vary. In addition, while the tax policies towards the elderly have changed frequently over time in a number of states, in other states such policies remained relatively stable (Conway and Rork 2008). In this paper we exploit the changes in state income tax preferences over time, including exemptions, deductions, and credits, to explore the impact such tax policies have on the living arrangements and welfare of the elderly population. In the existing literature, the living arrangements of the elderly are outcomes of joint decision making between the old-age parents and their children. While earlier research work explained elderly's decision of independent living as a result of their income increase and the children's preferences (Kotlikoff and Morris 1988), more recent work have adopted an intra-family bargaining framework that focuses on the strategic interactions between the children and the parents (Konrad et al. 2002; Pezzin and Schone 1999; Pezzin, Pollak, and Schone 2006). However, despite their importance and obvious policy relevance, studies that analyze the impact of public policies on elderly living arrangements are scarce. This paper advances the existing literature by analyzing the effect of state income tax preferences on the elderly's living arrangements and welfare. As tax policies become more favorable towards the elderly, their options of living independently become higher in value, which yields two testable implications. First, elderly individuals may exercise their option of living independently under favorable tax policies. Second, since tax preferences increases an elderly individual's wealth, and consequently bargaining power in the household, such preferential policies may have an indirect impact on elderly's welfare, especially for those who do not choose to live independently. Using the Panel Study of Income Dynamics (PSID), we construct cohorts of households whose heading member reaches the age of 65 at each wave of the survey. The PSID is conducted annually starting from 1968 and contains a nationally representative sample of all ages and allows us to control for such factors as health outcomes, employment, and family composition changes. Combining our cohorts with the newly constructed dataset on type and evolution of state income tax preferences for the elderly developed by Conway and Rork (2008), we then empirically test the implications of our model.