2
By examining important cases that do not readily fit into either category, this paper
problematizes the distinction between deconcentration and devolution. It does not question
the importance of subnational elections as signal events that alter the calculations of would-
be decentralizers at the center. In Latin America and other regions, the spread of
subnational elections in the last two decades has set in motion deep changes in the political
careers that politicians pursue, the lobbying strategies that interest groups adopt, and the
demands that are now emerging from subnational governments. Indeed, in many countries,
using electoral mechanisms rather than appointment procedures to constitute subnational
governments is one of the most significant features that distinguishes democracy’s third
wave from its earlier waves. Even if subnational elections are imperfect, and even if national
party leaders continue to exert some degree of influence over subnational incumbents and
candidates, we should still expect these elections to alter the meaning of reforms that shift
power downward.
2
Placing too much emphasis on elections, however, leads us to overlook reforms that
expand the institutional capacity of subnational officials who are not elected, but who
nevertheless cannot be described accurately as agents of central government bureaucracies.
Salient examples of such reforms include changes that give un-elected subnational
governments control over the provision of schools and hospitals, the right to operate their
own banks and state-owned enterprises, and additional taxing authority. The simple
distinction between deconcentration and devolution cannot account for these changes
because they transcend the mere empowerment of central bureaucrats located below the
national level, but do not meet the higher standard set by separate elections for subnational
2
In O’Neill’s work, it is precisely the political independence of subnational governments from the center that
leads national party leaders to endorse devolution when they deem their prospects of holding onto the national
government to be sufficiently slim (O’Neill 1999 and 2003).