Portal de Periódicos da CAPES

Tax Competition with Heterogeneous Firms

2014; Routledge; Volume: 9; Issue: 3 Linguagem: Inglês

10.1080/17421772.2014.930164

1742-1780

Richard Baldwin, Toshihiro Okubo,

Fiscal Policy and Economic Growth

AbstractThis paper studies tax competition in an economic geography model that allows for agglomeration economies with trade costs and heterogeneous firms. We find that the Nash equilibrium involves a large country charging a higher tax than a small nation. Lower trade costs lead to an intensification of competition, a drop in Nash tax rates and a narrowing of the gap. Since large, productive firms are naturally more sensitive to tax differences in our model, large firms are the crux of tax competition in our model. This also means that tax competition has consequences for the average productivity of big and small nations' industries; by lowering tax rates, a small nation can attract high-productivity firms.Résuméla présente communication se penche sur la concurrence fiscale dans un modèle de géographie économique prévoyant des économies d'agglomération avec des coûts commerciaux et des entreprises hétérogènes. Nous relevons que l'équilibre de Nash comporte l'existence d'un taux d'imposition plus élevé dans les pays de grande taille que dans les petites nations, ce taux étant insuffisant sur le plan social. La réduction des coûts commerciaux engendre une intensification de la concurrence, une diminution des taux d'imposition de Nash, et une réduction de l'écart. Étant donné que, dans notre modèle, les entreprises productives de grande taille sont naturellement plus sensibles aux divergences fiscales, ces mêmes entreprises de grande taille constituent la question de fond de la concurrence fiscale dans notre modèle. Ceci signifie également que la concurrence fiscale comporte des conséquences pour la productivité moyenne de l'industrie dans les grands et les petits pays; en réduisant les taux d'imposition, le petit pays est en mesure d'attirer des entreprises à productivité élevée.Resumeneste estudio analiza la competencia fiscal en un modelo de geografía económica que contempla a las economías de aglomeración con costes comerciales y las firmas heterogéneas. Vemos que el equilibrio de Nash comporta al país grande que carga un impuesto más alto que el de la nación pequeña, con esta tasa excesivamente baja desde un punto de vista social. Los costes comerciales menores conducen a la intensificación de la competencia, a la reducción de las tasas fiscales de Nash y a una reducción de las diferencias. Debido a que las grandes firmas productivas son por naturaleza más sensibles a las diferencias fiscales en nuestro modelo, las grandes firmas son el punto crucial de la competencia fiscal en nuestro modelo. Esto también significa que la competencia fiscal afecta a la productividad media de la industria de naciones grandes y pequeñas; a través de la reducción de las tasas fiscales, las naciones pequeñas pueden atraer firmas altamente productivas.摘 要本文通过一个经济地理学模型对税收竞争进行了研究,该模型允许存在具有贸易成本和异质性企业的聚集经济。我们发现,在纳什均衡中,大国的税收比小国 高,但从社会角度上看,这一税率却又过低。较低的贸易成本会导致竞争加剧,纳什税率下降,差距缩小。在我们的模型中,由于生产力高的大型企业一般会对税收 差别更为敏感,因此,大型企业是税收竞争的关键所在。这还意味着,税收竞争会对大国和小国内的产业的平均生产力造成影响;通过调低税率,小国可以吸引高生产力的企业。Keywords: Firm heterogeneityspatial sortingNash equilibrium taxtax cooperationaverage productivityJEL classification: H32P16 We would like to thank Rikard Forslid, Gianmarco Ottaviano and Masahisa Fujita for their helpful comments and suggestions as well as participants in seminars at Tohoku University and Kobe University for their comments.This research is partially funded by Grant-in-Aid for Scientific Research (Japan Society for the Promotion of Science (JSPS)).We would like to thank Rikard Forslid, Gianmarco Ottaviano and Masahisa Fujita for their helpful comments and suggestions as well as participants in seminars at Tohoku University and Kobe University for their comments.This research is partially funded by Grant-in-Aid for Scientific Research (Japan Society for the Promotion of Science (JSPS)).Notes1. For instance, Ludema & Wooton (Citation2000), Kind et al. (Citation2000), Trionfetti (Citation2001), Forslid & Midelfart Knarvik (Citation2001), Ottaviano & Van Ypersele (Citation2005), Andersson & Forslid (Citation2003) and Baldwin & Krugman (Citation2004).2. Baldwin & Okubo (Citation2014) propose a generalized model of Baldwin & Okubo (Citation2006) to highlight the difference from the HFT model. According to the paper, it is not possible to solve equations due to over-identification when we assume free entry and exit as well as free relocation simultaneously. In a special case of free trade costs, it is solvable both simultaneously, as in Davies & Eckel (Citation2010).3. The model of Baldwin & Okubo (Citation2006) is based on the simplest economic geography model. Some full features as in the standard economic geography model are added by subsequent works. Okubo et al. (Citation2010) extends to the linear demand model. By adding intermediate inputs, Okubo (Citation2009) includes forward and backward linkages as in the standard economic geography.4. Since we take the range of varieties to be continuous, we speak of the 'mass' of firms with a particular marginal cost. We assume that the mass is the same for every level of marginal cost [this is demonstrated in Melitz (Citation2003) as the outcome of an endogenous entry/exit process].5. Relocation is always one-way from South to North. The least productive firms in North never relocate to South. Baldwin & Okubo (Citation2006) prove the relocation tendencies in a formal manner. See Baldwin & Okubo (Citation2006) for more details.6. This simplification uses mill pricing and cancels the (1–1/σ) terms.7. Since firms are atomistic, the first firm to move has no impact on the Δs.8. This result is known as the 'home market effect' in international trade.9. Note that the delocation process tends to raise the average efficiency of industry in the big region while it lowers the average efficiency in the South.10. The parameter values chosen for the simulations that generated the figure are σ = 2, ρ = 2, ϕ = 0.75 and s = 0.6.11. For convenience sake, we define 12. Note that T>T* from Result 2.13. The parameter values in Figure are σ = 2, ρ = 2, ϕ = 0.75 and s = 0.6.