5. Conclusions and suggestions for further research
In this paper we have set up a probabilistic voting model to explore the hypothesis
that tax competition improves public sector efficiency and social welfare when a
political distortion favours public sector employees. In our model the political
distortion induces politicians to create rents through high wages to public sector
workers in the absence of tax base mobility. If tax competition is introduced via
the lifting of capital controls, it will reduce the rents to public sector workers and
may well destroy them completely when the number of competing jurisdictions
becomes sufficiently large. However, tax competition will also cause an under-
provision of public goods by increasing the marginal cost of public funds. Our
analysis indicated that, in the presence of a political distortion favouring public
sector workers, a modest degree of tax competition involving a relatively low tax
base elasticity is likely to be welfare-improving, whereas unfettered tax compet-
ition among small jurisdictions is likely to be welfare-reducing, compared to a
hypothetical situation without tax base mobility. In particular, if tax competition
is sufficiently strong to eliminate all rents, a coordinated rise in capital taxation
will always be welfare-improving by offsetting the underprovision of public goods.
We also found that it may be welfare-enhancing to carry tax coordination beyond
the point where rents to public sector workers start to emerge.
Overall our analysis suggests that while the advocates of tax competition are
right in claiming that tax base mobility serves to reduce rent-seeking, it is a double-
edged sword that also tends to distort the supply of public goods, as argued by
supporters of tax coordination. Up to a certain point tax competition may play
a useful efficiency-enhancing role, but if it becomes too intense it is likely to be
welfare-reducing. Indeed, in a calibrated version of our model we were able to
identify an optimal intensity of tax competition, measured by the elasticity of
the tax base with respect to the tax rate. Our quantitative analysis suggested
that even very large political distortions can only justify a modest intensity of tax
competition. In our model tax competition thus seems a poorly targeted means
of curbing rents, compared to domestic institutional reform.
A natural extension of our analysis would be to allow for taxes on immobile
factors. One could then study whether tax competition for mobile factors will
reduce rents even if politicians can compensate for a lower revenue from the mobile
tax base by raising taxes on the immobile base. We believe the answer to this
29