Roger Bruce Myerson : Democratic decentralization and economic development
By Roger B. Myerson - keynote speaker for the introductory plenary of the World Summit of Local and Regional Leaders.
My thesis is that the relationship between local and national politics is vital for economic and political development, and that local democracy may be severely under-supplied in much of the world. I will argue that democratic development must begin by increasing the supply of leaders with reputations for using public funds responsibly to provide public services (and not just to give jobs to their supporters). Local democracy can help increase this supply of democratic leadership, but established national leaders may have vested interests in limiting it.
Analysis of economic development is incomplete without considering the structures of local political leadership. In history, we can find roots of underdevelopment in feudal systems of local governance. When first establishing rule in India (and after the 1857 mutiny), the British granted local power and privileges to local agents called zamindars. Holding local power as a permanent property right, the zamindars became a class of local leaders with a vested interest in the regime. The effectiveness of their feudal power proved remarkably durable. Long after independence, Abhijit Banerjee and Lakshmi Iyer (2005) find lower agricultural productivity and higher infant mortality in zamindar regions. The lesson is that feudalism can help to establish a stable political regime, but it can also have serious long-term economics costs. We must ask, how much of global underdevelopment has resulted from such strategies of traditional and colonial state-building?
Political systems differ crucially on whether governors and other powerful local officials are centrally appointed or locally elected. Either constitutional system can become self-sustaining.
Economic agency theory teaches us that such powerful offices be associated with large rewards that economists call large moral-hazard rents. Where governors are centrally appointed, these offices become valuable patronage prizes that national leaders offer to key supporters, and then national leaders cannot easily disappoint their supporters' hopes for such rewards.
On the other hand, where governors are locally elected, a move to centralize these offices would threaten the established positions of powerful the local power-brokers who hold these offices. Then national leaders may find that a reputation for working with local leaders within the existing constitution becomes an essential asset for building strong coalitions.
With central appointment, local officials depend on their relationships with national leaders and do not need trust of the local population. Holding these officials more accountable to local complaints would reduce the value of their offices as rewards for national political service.
Thus, centralization can degrade public services and weaken the state outside the capital, as observed by Robert Bates (2008).
Success of democracy requires more than just elections; it requires candidates who can actively compete to better serve the public interest. Voters would not reject a corrupt incumbent unless they could expect better from another candidate. Thus, the key to democratic development is to increase the supply of leaders who have good reputations for using public funds to provide public services, and not just to give patronage rewards to their active supporters. This supply of good democratic leadership can best develop when democracy is applied in both national and local governments. Elected local leaders who successfully provide better local public services can demonstrate qualifications to become strong candidates for higher office.
Thus, democratic local government can make national politics more competitive, lowering barriers against new entrants into national politics. (See Myerson, 2006.) But established leaders do not want more competition, and so national leaders may generally prefer a more centralized political system.
On the other hand, national democracy can strengthen local democratic competition, as national parties can support alternatives to established local bosses. Thus, strong competitive democracy depends on both a multi-party national assembly and elected local councils with autonomous local responsibilities.
This perspective has important implications for foreign development assistance programs. For positive political development, the essential measure of a assistance project's success is how it enhances reputations of the political leaders who direct it. Outputs of public goods count only towards this political end. So all public services should be directed by indigenous leaders, but they should include both national and autonomous local leaders. To cultivate leadership at all levels, assistance should be distributed to national and local governments (perhaps also to minority parties). Largest share of assistance funds should go to national government, but conditioned on some decentralization also to local public service agencies, as has been argued by Paul Collier (2007). Foreign donors should require transparent public accounting for all funds, both by the donors and by the national finance ministry. This accounting should be to the local population, so they can learn what their leaders have spent, and what this spending has achieved.
- Abhijit Banerjee and Lakshmi Iyer, "History, institutions, and economic performance," American Economic Review 95(4):1190-1213 (2005).
- Robert Bates, When Things Fell Apart (Cambridge U. Press, 2008).
- Paul Collier, The Bottom Billion (Oxford U. Press. 2007).
- Roger Myerson, "Federalism and incentives for success of democracy," Quarterly Journal of Political Science 1(1):3-23 (2006).
The full paper is available at : The University of Chicago