Agglomeration, integration and tax harmonisation

https://doi.org/10.1016/S0014-2921(02)00318-5Get rights and content

Abstract

Consideration of agglomeration reverses standard theoretical propositions in international tax competition. We show greater economic integration may lead to a ‘race to the top’ rather than a race to the bottom. Also, ‘split the difference’ tax harmonisation may harm both nations, a result that may explain why real-world tax harmonisation is rare. The key is that industrial concentration creates ‘agglomeration rent.’ The ‘core’ region can thus charge a higher tax rate without losing capital. The size of such rent is a bell-shaped function of the level of integration, so the tax gap first widens before narrowing as integration increases.

Introduction

Does close economic integration, especially in the face of the growing mobility of capital both physical and human, require harmonisation of tax rates? Many observers believe that it does. It is often argued that the nations of the European Union, in particular, must agree on common tax rates if they are to avoid a “race to the bottom” that will undermine their relatively generous welfare states. The logic seems straightforward: other things being equal, producers will move to whichever country has the lowest tax rates, and absent any coordination of tax-setting the attempt to attract or hold on to employment will lead to a competition that drives tax rates ever lower.

But things are not necessarily equal. Countries with generous welfare states tend to be countries that have long been wealthy; such nations offer capital the advantages of an established base of infrastructure, accumulated experience, etc. – in short, they offer favourable external economies. And within limits this presumably allows them to hold on to mobile factors of production even while levying higher tax rates than less advanced nations. On the other hand, should the tax rate get too high, the results could be catastrophic: not only will capital move abroad, but because that movement undermines agglomeration economies it may be irreversible.

What this suggests is that in the face of the sort of agglomerative forces emphasised by the “new economic geography”, the tax game played in the absence of harmonisation may be something subtler than a simple race to the bottom. Advanced countries may be more like limit-pricing monopolists than Bertrand competitors; their interaction with less advanced countries need not lead to falling tax rates, and might well be consistent with the maintenance of large welfare states.

The purpose of this paper is to think about international tax competition and harmonisation in the presence of significant agglomeration economies and goods market integration. The existing literature in this area is limited. Most of the vast tax-competition literature – see the survey by Wilson (1999) for instance – works with the ‘basic tax competition model’ (BTCM). This is a one-period model featuring a single good produced by two factors; labour, which is immobile between regions and capital, which is mobile. Trade costs are zero, firms face perfect competition and constant returns, so there is no trade among regions and capital faces smoothly diminishing returns. Typically, governments chose the capital tax rate (the labour tax rate is either ignored or assumed to be identical to capital's) in a Nash game. The standard approach is to compare equilibrium tax rates with no capital mobility and with perfect capital mobility; or to compare non-cooperative with cooperative tax setting both under prefect capital mobility. The customary result – equilibrium taxes are sub-optimally low – has been greatly extended and modified, but still remains the received wisdom on tax competition among social welfare maximizing governments. In one extension, where governments are assumed to deviate from social welfare maximisation, tax competition may improve welfare by moving equilibrium rates closer to the social optimum (in a typical second-best fashion). Two aspects of this literature are noteworthy. First, an analysis of tighter goods market integration and tax competition is absent since the focus is on heightened capital mobility. Second, although a small branch of this literature (e.g. Janeba, 1998) does consider imperfectly competitive firms, the standard tax-competition literature entirely ignores issues of agglomeration externalities.

Baldwin et al. (2003) review the tax and agglomeration literature that has emerged since the 1998 draft of our paper in detail, but three papers are particularly noteworthy. The first paper in this area is Ludema and Wooton (2000). This paper studies the impact of varying both factor-mobility costs and trade costs and seems to find that lowering either cost may – in contrast to the standard BTCM result – result in higher taxes being chosen in a tax competition game among governments. These authors, together with Andersson and Forslid (1999), and Kind et al. (1998) make the important point that agglomeration creates rents for the mobile factor that can be taxed. This point also plays an important role in the analysis below. Our paper focuses solely on the case where industry is fully agglomerated in one region and we find that the tax gap between nations is bell-shaped in trade openness – first rising and then falling as trade gets more open. Most importantly, we explicitly consider the implications of agglomeration forces for tax harmonisation schemes.

The paper begins by briefly surveying the standard tax-competition literature's main results. Then we present some empirical trends in taxation within Europe. Next we turn to a simple, stylised model of economic geography in the face of taxes. We note that our main results turn on properties that hold in a wide range of economic geography models, so we conjecture that our results would hold in many models, but to be specific, we work with a model that is simple enough to allow us to obtain analytic results.2 In this subsequent section, our specific model serves as a basis for examining the game that uncoordinated tax authorities might play. The final section presents concluding remarks.

Section snippets

Standard tax competition results

We briefly discuss the main tax competition results and their logic to boost intuition for why the inclusion of agglomeration forces leads to such different results.

Taxation in Europe: stylised facts

Increased economic integration is not a new development. Within Europe, in particular, barriers to trade both natural and artificial have been falling more or less continuously since the late 1940s. So it is possible, by looking at previous European experience, to get some idea of how increased integration and tax competition among nations have interacted in the past.

In making these comparisons we think of Europe as being divided into two parts: an advanced “core” that benefits from the

Tax competition with trade and factor mobility

We present the model of the underlying economy before turning to the tax competition game. As shall become clear below, the key to our argument is the existence of agglomeration rents. Baldwin et al. (2003) show that such rents arise in a wide range of economic geography models, including those of Krugman (1991) and Venables (1996). We thus conjecture that our results would hold in a broad range of models. To be concrete, and to be able to get analytic results, the model we adopt is a solvable

The tax game

The tax competition literature assumes that governments value tax revenue for one of two reasons. If the government is benevolent, tax revenue is used to finance public goods and the government cares about revenue since consumers like such goods. If the government is modelled as a ‘Leviathan’, i.e. it does not maximise social welfare, the government's objective is either to maximise the size of the state or to maximise its own utility, which may in turn depend on the probability of its

Tax harmonisation

As it turns out, this setup suggests that tax harmonisation has somewhat unexpected results. In the basic tax competition model, tax harmonisation basically entails a shift from a non-cooperative tax game to a cooperative tax game; Pareto improvement from the government's perspective follows by definition. In stark contrast, harmonisation makes one or both countries worse off when agglomeration forces are present.

To see this, consider first the most straightforward tax harmonisation scheme,

Concluding remarks

This paper looks at the impact of tighter goods market integration on international tax harmonisation and tax competition when agglomeration economies are significant. The presence of agglomeration forces makes the economy “lumpy” in the sense that industry tends to stay together, either all in one region or all in the other. The lumpiness also gives industrialised nations – the so-called core nations – an advantage over the less industrialised nations – the so-called periphery. Agglomeration

Acknowledgements

We thank Federica Sbergami and Tommaso Mancini for excellent assistance and the Swiss National Science Foundation for financial support.

References (13)

There are more references available in the full text version of this article.

Cited by (383)

  • Spatial growth theory: Optimality and spatial heterogeneity

    2023, Journal of Economic Dynamics and Control
View all citing articles on Scopus
1

This paper was written while Baldwin was visiting MIT in 1998/1999, with the first draft in December 1998 and revisions in June 2000 and April 2002.

View full text