Capital Taxes, Labor Taxes and the Household
A recent literature has found that married couples share risks only partially. When the labour income of one spouse decreases (e.g. due to an unemployment shock), the couple allocates a lower weight to her well-being. This implies that this individual’s consumption absorbs the largest fraction of the overall drop in the family’s resources.
In our paper “Capital Taxes, Labor Taxes and the Household” we use a dynamic life cycle model which accounts for heterogeneity in wealth, productivity and gender, to investigate how the limited risk sharing within the household impacts individuals’ preferred structure of capital and labour taxes. There are several reasons why the interplay between taxes and the intrahousehold allocation is of particular interest. First, most of the macroeconomics literature on the optimal tax code considers the effects of policies on the inequality between households; the effects of taxes on the within- households distribution of resources is typically overlooked. Second, whereas labour income is a resource which is privately held in the household (giving rise to partial risk sharing), wealth is a commonly held resource. This observation, which can be justified empirically, implies that when the household is wealthier the limited risk sharing problem becomes less severe, due to the fact that labour income is a less significant fraction of total resources. In the paper we formalize this intuition, showing that indeed higher wealth improves the risk sharing possibilities of households.
We then ask: ‘How does the mixture between capital and labour taxes affect the intrahousehold allocation and welfare?’ With our quantitative model we consider a policy reform which abruptly eliminates capital taxation from the economy and finances government spending through increased labour taxes. This policy has been considered by many papers in the literature, and therefore offers a very useful benchmark. Our finding is that though this policy change indeed improves risk sharing within the household, it exerts only a moderate effect on this margin. The reason can be traced to the life cycle pattern of savings produced by the model and which mimics the data closely. Young couples, who could hypothetically gain more from the policy change, hold very little wealth and their savings do not respond strongly to the shift in capital taxes. Young individuals typically wish to borrow against their future income and an increase in after tax returns will not induce them to increase their savings by much. Older households hold substantially more wealth (e.g. to finance retirement) but for them the partial risk sharing is less severe to begin with; they also do not gain much from the shift in policy.
A considerable literature in quantitative macroeconomics has studied optimal capital and labour taxes assuming that households are ‘single individuals’, de facto leaving aside all aspects of the intrahousehold allocation. Our findings however, affirm the conclusions from this literature, when we explicitly consider multimember households we identify only marginally significant effects. We are however convinced that the model we proposed can be useful to evaluate policies which exert a direct impact on the bargaining power of individuals and therefore hold promise to influence profoundly the intrahousehold allocation and risk sharing. Such policies are, for example, changes in the progressivity of taxes, or tax reforms which enable spouses to file separately for taxes.
Read the full article Capital Taxes, Labor Taxes and the Household published in Journal of Demographic Economics