Agglomeration and tax competition

https://doi.org/10.1016/j.euroecorev.2005.01.006Get rights and content

Abstract

Tax competition may be different in ‘new economic geography settings’ compared to standard tax competition models. If the mobile factor is completely agglomerated in one region, it earns an agglomeration rent which can be taxed. Closer integration first results in a ‘race to the top’ in taxes before leading to a ‘race to the bottom’. We reexamine these issues in a model that produces stable equilibria with partial agglomeration in addition to the core–periphery equilibria. A bell-shaped tax differential also arises in our model. Therefore, the ‘race to the top’ result generalises to a framework with partial agglomeration.

Introduction

Conventional wisdom holds that capital tax competition leads to a ‘race to the bottom’. Under the assumptions of constant returns to scale and perfect competition, the standard model shows that the desire to attract physical capital leads to inefficiently low tax rates.1 In this framework, deeper economic integration would lead to fiercer competition and thus falling tax rates.

Things are different in models of the ‘new economic geography’.2 In the ‘core–periphery’ model, under certain parameter ranges, industry is completely agglomerated in one region. An agglomeration rent accrues to mobile capital in the core region. Therefore, taxing capital in the core region will not lead to an outflow of capital as long as the tax gap is smaller than the agglomeration rent. Moreover, since this agglomeration rent is a bell-shaped function of the level of trade integration, the tax gap is also bell-shaped. Hence, in contrast to standard international tax-competition results, closer integration may first result in a ‘race to the top’ before leading to a ‘race to the bottom’ (Baldwin and Krugman, 2004). Baldwin and Krugman (2004) suggest that this may explain the observed pattern of the evolution of corporate taxes in the European Union.

All of the theoretical papers on agglomeration and tax competition to date use the core–periphery model or a variant thereof (see Ludema and Wooton, 2000, Kind et al., 2000, Andersson and Forslid, 2003, Baldwin and Krugman, 2004, Baldwin et al., 2003). In these models, stable equilibria are either those where industry is divided symmetrically or where all of industry locates in one of two countries. This feature is arguably extreme and not very realistic (see, e.g., Ottaviano and Thisse, 2004). By focusing on the standard core–periphery model or on very close descendants, the previous literature has left out an important class of locational equilibria. Ludema and Wooton who have themselves contributed to the development of models containing stable equilibria with partial agglomeration (Ludema and Wooton, 1999) are fully aware of this omission. However, they note that an analysis of such cases would become complex and therefore abstain from analysing this case in their tax competition paper (Ludema and Wooton, 2000, p. 342).

The present paper also analyses tax competition with agglomeration forces. But we depart from the literature by using an economic geography model which, in addition to the core–periphery equilibria, allows for stable locational equilibria with only partial agglomeration of firms in one of two countries. An interesting question is how far the results of the core–periphery model – in particular the possible ‘race to the top’ – generalise to such a framework with partial agglomeration.

To make the problem tractable, the present paper makes a few strategic simplifications. First, we use a simple model out of the class of models which feature stable partial agglomeration equilibria (the model of Pflüger, 2004). Second, we follow Ludema and Wooton (2000) by considering lump sum taxes only. Third, to sharpen the tax game, we follow the standard tax competition literature in assuming that taxes are levied on the mobile factor only (Zodrow and Mieszkowski, 1986).3 Fourth, we adopt the reduced-form government objective function of Baldwin and Krugman (2004) and stick to their simple quadratic approximation of this objective function.

All these simplifications notwithstanding, we have to rely on numerical simulations at some stages of our analysis. However, our reconsideration of the tax game yields the results that the tax shield provided by agglomeration forces is more general than suggested in the previous literature. In particular, we show that a tax differential may arise as an equilibrium of the tax game even when there is only partial agglomeration and the mobile factor does not derive an agglomeration rent. We also find that the tax gap is a bell-shaped function of trade costs: the ‘race to the top’ result of the core–periphery model does indeed generalise to a framework with partial agglomeration. This reinforces the plausibility of the hypothesis that agglomeration forces account for observed (corporate) tax differentials in integrating regions such as the European Union (see Baldwin and Krugman, 2004).

Stable equilibria with partial agglomeration have been shown to emerge under a variety of circumstances but do not exist in the standard core–periphery model. The clue to their existence lies in the fact that they are obtained in models that enrich the standard core–periphery model by incorporating further centrifugal forces or by weakening centripetal forces (Pflüger, 2004, Ottaviano and Thisse, 2004). Helpman (1998) incorporates housing (a non-traded good) in the model. At low trade costs housing rents act as a force that disperses the mobile factor. Fujita et al. (1999, Chapter 18) introduce congestion costs in a general way, and obtain a result similar to Helpman. Puga (1999) and Fujita et al. (1999, Chapter 14) allow for decreasing rather than constant returns to labour in the production of the agricultural good. Under this assumption, the manufacturing sector does not face a horizontal labour supply curve as in the standard core–periphery model but rather an upward sloping curve which acts as a force of dispersion. Ludema and Wooton (1999) assume limited factor mobility. If the mobile factor does not necessarily move in response to any (even marginal) utility differential, and if the mobile population has mobility costs which increase in the distance from the mean, stable equilibria with partial agglomeration obtain at low trade costs. Another route to obtain such equilibria is by changing the upper-tier utility function (Pflüger, 2004). If the standard Cobb–Douglas upper-tier utility is replaced by a logarithmic quasi-linear utility function, the demand and supply linkages of the core–periphery model are retained. However, the removal of income effects from the demand for manufactured goods weakens the demand linkage and again allows equilibria with partial agglomeration to be stable. We use this model in our analysis of tax competition with agglomeration forces. It should be noted that all of these modifications of the core–periphery model retain equilibria with full agglomeration for certain ranges of trade costs.

The structure of the paper is as follows. The next section introduces the model. Section 3 characterises its three types of stable equilibria without taxes (the ‘no-tax equilibria’). Section 4 introduces the reduced-form government objective function and discusses the nature of equilibria with taxes. The tax game is taken up in the two subsequent subsections. We begin by analysing tax competition in the case of a core–periphery setting. We then undertake an analysis of the tax game when partial agglomeration is a stable locational equilibrium. Section 5 concludes.

Section snippets

The model

Our theoretical analysis draws on Pflüger (2004), who develops a model that gives rise to stable equilibria with partial agglomeration of firms. The model deviates from the standard core–periphery model in two respects. First, as in Forslid (1999) and Forslid and Ottaviano (2003), it is assumed that the fixed cost in the manufacturing sector consists of a separate internationally mobile factor. This makes the core–periphery model analytically solvable without changing its basic features.

No-tax equilibria

This section characterises the equilibria in the model without taxes, i.e., where t=t*=0. The long-run location decision of entrepreneurs is governed by the (indirect) utility differential VK-VK*=αln(P*/P)+(R-R*), which arises for a given allocation of the capital stock in the short run.6 Let KW

Taxes and the government objective function

We now turn to the analysis of taxes and tax competition. For simplicity we normalise the endowments of factors such that K=K*=1, and hence, KW=2, and ρ=ρ*=1. Under these assumptions, the international utility differential of the mobile factor is given byΩ(λ)-(t-t*),where Ω(λ) is as defined in (10) and the other arguments have been suppressed.

Note that Ω(λ) does not depend on tax rates. Hence, equilibria can be studied by superimposing Ω(λ) with the tax differential t-t*. For the case of

Conclusion

In this paper we have reconsidered the analysis of tax competition under agglomeration forces. Whereas the previous literature has considered either symmetric equilibria or location equilibria with complete agglomeration, we have presented a model that also allows for partial agglomeration. We have shown that in the case with partial agglomeration, the partial core can maintain a positive tax gap even though no agglomeration rent accrues to the mobile factor. An interesting difference to

Acknowledgements

Paper presented at the Royal Economic Society annual meeting in Swansea, the European Economic Association meeting in Madrid, the IIPF meeting in Milan and seminars in Konstanz and Mainz. We would like to thank Pio Baake, Richard Baldwin, Andreas Haufler, Jochen Michaelis and two referees for helpful comments. The first author is grateful for financial support from the German Science Foundation (DFG) through SPP 1142 ‘Institutional design of federal systems’.

References (23)

  • D. Puga

    The rise and fall of regional inequalities

    European Economic Review

    (1999)
  • Cited by (105)

    • Institutional quality and FDI location: A threshold model

      2022, Economic Modelling
      Citation Excerpt :

      The remainder of this paper is as follows: Section 2 introduces a basic FDI location model and develops an empirical panel institutional quality threshold model for FDI location decisions; Section 3 provides data sources and stylized facts; Section 4 provides the empirical results and robustness tests; and Section 5 summarizes and presents policy implications. We adopt a geography model for determining FDI location decisions on a firm's regional competition in an oligopolistic industry, following Haufler and Wooton (2010), Baldwin and Krugman (2004), and Borck and Pflüger (2006). However, in contrast to those studies, this study explores the role of institutional quality in examining the relationship between labor costs and FDI location decisions in the model.

    • Technology and tax systems

      2020, Journal of Public Economics
    • Fiscal equalization and the tax structure

      2020, Regional Science and Urban Economics
    View all citing articles on Scopus
    View full text