Local Nonproperty Revenues
In this paper, Jorge Martinez-Vazquez focuses on the current theory and practice of non-property tax assignments to local urban governments in developing and developed countries. Although the theoretical rationale is clear for the presence of a basket of tax instruments, the international experience in both developing and developed countries is mixed and uneven. While a few urban governments have introduced economically attractive tax sources other than the property tax, a vast majority of urban governments in developing countries, and also in many developed countries, still struggle with the imperative of revenue adequacy to cover their growing expenditure needs and obligations. This paper first reviews the theoretical foundations for the use of taxes other than the property tax in local governments. Next, actual practices in tax assignments are surveyed; special attention is paid to political economy and institutional capacity issues as potential drivers for the observed differences between an optimal metropolitan revenue system and those that are actually observed.
The paper concludes that effective fiscal decentralization requires meaningful levels of revenue autonomy at the subnational level. Efficient spending decisions at the local level require that decision makers face the true marginal cost of funds. Besides providing revenue sufficiency, tax autonomy brings political accountability and higher fiscal responsibility at the subnational level. Meaningful subnational revenue autonomy typically requires the cohabitation of productive tax bases with the central government and discretion to set tax rates for taxes selected from a closed list. Subnational administration of subnational taxes can be desirable to enhance accountability, but some flexibility is desirable to allow centralized tax administration to take advantage of economies of scale in management and information. There is ample consensus that user charges and fees are the most appropriate source of revenue for local governments, fitting almost perfectly within the benefit principle. Nevertheless, it is not generally feasible to finance all local services with user charges. Better choices of local taxes include property taxes and betterment levies, vehicle and transportation taxes, local business taxes, flat (piggyback) individual income taxes, and final sales and excise taxes.
Two puzzles remain in the practice of revenue assignments that require additional research. Although there is a role for historical factors such as colonial roots, it remains difficult to explain why inappropriate assignments and bad design have proved so difficult to reform over the years in so many countries. A separate puzzle has to do not with their design but their actual implementation. Often the revenue autonomy provided in the revenue assignments goes unused by the same subnational governments demanding additional funding from their central governments. Future research should pay closer attention to the political economy of revenue assignments.
This paper was presented at a 2011 conference at The Brookings Institution organized by the Lincoln Institute of Land Policy and is Chapter 8 of the book Financing Metropolitan Governments in Developing Countries.