The literature on state capacity and development has attracted renewed interest over the last few years, in particular in the study of violent conflict. Yet, state capacity is conceived differently depending on whether the interest lies in the state’s power to discourage violent conflict, in its ability to administer efficiently, or simply in its capacity to foster economic development. In this article, we examine the conceptual and empirical links between state capacity and bureaucratic autonomy and discuss the conditions under which these converge or differ. By making use of panel data from 1990-2010, we then estimate the separate effect of state capacity and bureaucratic autonomy on two of the Millennium Development Goals indicators: child mortality and the prevalence of tuberculosis. We present evidence suggesting that a) bureaucratic autonomy has a stronger impact than commonly used measures of state capacity; and b) that both bureaucratic autonomy and state capacity play a more important role for these indicators than traditional macroeconomic variables such as GDP growth.
