Economic Disadvantages of Decentralization:

Despite its advantages, a system of sub-national governments is likely to be ineffective in resolving certain policy issues. For example, if local governments are relied on to provide national public goods, the result is certain to be substantial inefficiency. Each community would be responsible for its own defense against foreign aggression, but all communities taken together would probably be under- defended. The reason is that a large part of the benefits from, for example, a missile system provided by one community accrue to residents of other communities. In determining how much to spend on missiles, each community would consider only the benefits its  residents receive and ignore the benefits to non-residents. The free rider phenomenon is relevant here, just as it is when we consider whether an individual has an incentive to contribute to the financing of a good that benefits other persons in addition to himself or  herself.

Because each community has an incentive to free ride, fewer resources would be devoted to defense than are justified by the interests of all persons in the nation considered together.

Efficient provision of a public good that benefits persons in all communities thus necessitates a central government. The key issue  here is the geographic area over which persons necessarily benefit from provision of the good. Some goods, such as a sewer system or police force, have benefits that extend over a limited area, and a city or country government can supply the good more efficiently because most, if not all, of the benefits and costs occur within one locality. Some environmental policies may have effects over larger regions and require state governments, but defense is clearly the province of the federal government. Thus, sub-national  governments will not be efficient in providing all types of government services. 

A second area in which sub-national governments will be relatively ineffective is in redistributing income. If one community embarks on a redistributive program by taxing its higher-income residents and transferring the proceeds to its poorer residents, two things will happen eventually to hinder the goal of providing assistance to the poor. First, wealthy residents can move to another community  where taxes are lower.

Second, poor persons in other area can move to the community providing higher welfare benefits. As a result, the average per capita income in the community will fall, and it will become increasingly difficult to finance high welfare benefits. Thus,  the very consumer  mobility that is beneficial when sub-national governments engage in nonresidistributive programmes tends to limit their use of  redistributive measures. This remains true even if those with higher incomes genuinely wish to help the poor. Each wealthy individual  who leaves a community has a negligible effect on the extent to which the poor are helped, so it is in each person's interest to leave. If all wealthy persons leave, of course, there can be no redistribution. Here is simply the free rider problem once again. Any significant degree of income redistribution must be carried out by a central government. Even in this case, there are limits because wealthy  persons can always leave the country, and poorer immigrants may be permitted to enter. Mobility, however, is far more restricted  between countries than between regions within a country, so a national government has greater latitude in carrying out redistributive programmes. Moreover, insofar as our concern is with poor persons, regardless of where they reside within the nation, only a national  program is capable of accommodating this goal by helping the poor in all regions. This does not mean that local governments cannot  adopt policies that benefit some residents at the expense of others, because they obviously do, but there are limits. In addition, the redistribution that occurs often takes a form that cannot be avoided by moving from the region. If a heavy property tax is placed on  land owned by a wealthy person, the landowner cannot fully avoid the burden by selling the land; the sale price will be reduced  because the buyer will pay less for heavily taxed property.

The immobility of property, especially land, makes it possible for local governments to engage in some redistribution, but still only to a limited degree. A third policy area in which sub-national governments are relatively inefficient is in the pursuit of macroeconomic  objects. Although our primary concern here is not macroeconomics, it should be pointed out that sub-national governments have little effect on employment and price levels within a specific region. These governments have no power to change the money supply, and any fiscal policies would be greatly diluted through the movement of persons and goods across jurisdictional boundaries. The national government must assume the responsibility for macroeconomic stabilization policy. The macroeconomic stabilization function will be better performed by central government than by local governments. Because of the openness of small local economics, the government expenditure multiplier will be low and fiscal policy will have less impact for stabilization purposes. In this case the effect of local fiscal expansion, for example, will fall on the 'imports' from other local economics. The fiscal multiplier in the expansionary locality will then be reduced. Any particular local government may be forced to incur an extremely large budget deficit in order to have an  expansionary effect on the local economy.