jurisdictions. The key point is that there is a trade-off between equity and accountability: if
resources are shared equally between sub-central governments, the incentive effects from
fiscal autonomy disappear. This has become one of the central issues that countries have
faced in deciding on the appropriate level of fiscal autonomy for sub-central tiers of
government.
The second argument against complete fiscal autonomy is that one has to avoid tax
externalities, in the interests of economic efficiency. ‘Tax exportation’ (McLure, 1967) is one
such problem. If taxation decisions by sub-central governments impinge on non-residents,
then local residents and politicians will not internalise the costs of public services and there
will be a degree of over-provision. In essence the same problems caused by grants re-emerge.
Of course, insofar as non-residents benefit from service provision, then there could be under-
provision, a problem that often arises with large metropolitan areas and satellite towns. The
way to deal with this problem is to ensure that different types of taxes are assigned to
different levels of government so as to avoid tax exportation between sub-national
jurisdictions. Indeed, this is a key argument for not devolving natural resource taxation. The
dangers arising from not centralising tax decisions on natural resources are two-fold: that it
would lead to tax exportation; and that resource-rich regions and central governments would
both seek to compete for the taxation revenues, leading to excessive taxation on the natural
resource. For example, suppose that a royalty tax on oil and gas were assigned to the sub-
central government, whilst the central government could also tax the same natural resource
through taxes on corporate profits. This leads to both tiers of government excessively taxing
the same base.
A third limit on fiscal autonomy is a consequence of the potential migration of factors of
production. Tax competition is a fact of life in most federal systems, and the mobility of
16