Income splitting for tax purposes results in more specialization of wives, but does this in turn generate more gender inequality? In my dynamic bargaining model with a divorce threatpoint, I find that who controls the couple's labour supply plays a crucial role in establishing this link. If spouses choose their labour supply non-cooperatively, only the husband's increase -but not her own decrease -in labour supply introduces a negative term in the wife's change in welfare. If the wife does not control her own labour supply, a decrease in her own labour supply introduces an additional negative term. JEL classification: H31, D13Fragmentation du revenu, spécialisation et répartition intra-familiale. La fragmentation du revenu aux fins d'impôt résulte en une plus grande spécialisation desépouses, mais est-ce que cela engendre davantage d'inégalité entre genres? Dans un modèle de marchandage dynamique avec menace de divorceà la clé, on découvre que la personne qui contrôle l'offre de travail du couple joue un rôle crucial dans la réponseà cette question.

The conventional wisdom in public …nance is that local governments should …nance public services, including those provided to businesses, with user charges that function as bene…t taxes, and should in particular avoid in-e¢ cient source-based taxes on highly mobile capital. However, if user charges are not available or infeasible, several public …nance experts have recently suggested that taxes on local production, such as an origin-based VAT, are a desirable alternative in that they serve as relatively e¢ cient "bene…t-related" taxes. We examine this contention formally in a model in which business public services must be …nanced with either a source-based tax on mobile capital, such as a property tax, or a tax on production, such as an origin-based VAT. In general, both a capital tax and a production tax are ine¢ cient. However, consistent with the "bene…t-related" view, the production tax is e¢ cient if the production function belongs to the knife-edge case between log sub-and log supermodularity with respect to capital and public services (e.g., a Cobb-Douglas production function), while the capital tax results in underprovision of public services in this case. Similarly, if the production function is log submodular with respect to capital and public services (e.g., a CES production function with substitution elasticity greater than one), a production tax is again less ine¢ cient than a capital tax, although both taxes result in underprovision of the public service. Finally, if the production function is log supermodular (e.g., a CES production function with substitution elasticity smaller than one), a production tax results in overprovision of the public service, while the e¤ects of a capital tax-and thus the relative e¢ ciency properties of the two taxes-are theoretically ambiguous.

No abstract

Although most of the tax competition literature focuses on the provision of local public services to households, several papers analyze tax competition when capital taxes are used to finance local public services provided to businesses, examining the conditions under which such services are provided efficiently, under-provided, or over-provided. In addition, several prominent observers have noted that "benefit-related" business taxation is desirable on both efficiency and equity grounds and argued that such taxation should take the form of a tax based on production, such as an origin-based value-added tax. We evaluate this contention in this paper, comparing the relative efficiency properties of these alternative business taxes. Our simulation results suggest that under many, but not all, circumstances it is more efficient to finance business public services with an origin-based production tax rather than a source-based capital tax.

While many theoretical works, particularly in family economics, rely on the transferable utility (TU) assumption, its exact implications in terms of individual preferences have never been fully worked out. In this paper, we provide a set of necessary and sufficient conditions for a group to satisfy the TU property. We express these conditions in terms of both individual indirect utilities and individual demand functions. Last, we describe the link between this question and a standard problem in consumer theory (initially raised by Gorman 1953), and explain why a similar characterization in terms of direct utilities cannot obtain.