This paper summarises the policy implications of a growing debate about the connections between taxation and the quality of governance in developing countries. Taxation - or the absence of tax - impacts on the quality of governance through two main channels. The first relates to the degree of dependence of governments on general taxation for their financial resources. Many governments do not need to make much tax effort because they have large non-tax incomes from oil, gas and mineral exports or from foreign aid. State elites are then financially independent of citizen-taxpayers. This changes the political incentives that they face, and the ways in which they seek to obtain, use and retain power. The long term consequences for governance are malign: state elites are less responsive and accountable to citizens; and, depending on the sources of non-tax revenue, may have less incentive to build up the political and organisational capacities of the state. States are likely to be simultaneously arbitrary and weak. All else being equal, the dependence of governments on general taxation has positive effects on the quality of governance. But that relationship is not automatic. How governments tax also matters. We cannot assume that, because they are fully dependent on taxation for revenues, governments will be capable, accountable, or responsive. They may levy taxes coercively, and thereby damage state-society relations and reinforce poor governance. Public authorities in contemporary poor countries face some incentives to tax coercively. Establishing more consensual taxation practices is an important practical route to improving governance. Aid donors could play a more constructive role.
Also published in Tax Notes International Vol 47(1): 79-98, July 2007.
To access a two-page summary of this paper, click this link.
IDS Working Paper No. 280, Brighton, UK: Institute of Development Studies, ISBN 978-1-85864-645-6, 39 pp.
Programme 3: State Capacity. How Does Taxation Affect the Quality of Governance? IDS Working Paper No. 280.