Allocation Without Need: Distributing a Special Autonomy Transfer Across Districts

Abstract

A special autonomy transfer is justified by the distinctive needs of the territory that receives it. Whether the transfer reaches the parts of that territory where need is greatest is a separate question, and one that is answerable from published records. This study tests it for a large subnational special autonomy scheme, using official allocation decisions for two budget years against official district indicators of poverty, child stunting and human development across twenty-three districts and cities. Four findings follow. Allocation is not statistically associated with any need indicator in either year: against poverty the correlation is negative but insignificant, against stunting close to zero, and against the human development index effectively nil. Allocation varies less than need does, with a coefficient of variation of 0.181 against 0.249 for poverty and 0.210 for stunting, so the instrument compresses rather than offsets the differences it addresses. Allocation is almost perfectly persistent between the two years, correlating at 0.967 with a near-uniform reduction averaging 11.7 per cent, which identifies the key as fixed rather than responsive. And the one district characteristic that does predict allocation is forest cover, a proxy for territorial area, at a rank correlation of 0.514. The district with the highest stunting prevalence, 47.9 per cent, ranks twenty-second of twenty-three; the district with the lowest poverty rate and the highest human development index ranks third. A needs index belongs in the statute rather than in an annual administrative decision.