The Flypaper is in the Composition, Not the Level: Evidence from South Sulawesi

The flypaper effect literature asks whether intergovernmental transfers stimulate local spending more strongly than own-source revenue. That question is almost always tested on total expenditure. This study moves the test to expenditure composition. The data cover 24 districts and cities in South Sulawesi for fiscal years 2015 to 2025, comprising 264 observations drawn from budget realization reports published by the Directorate General of Fiscal Balance. The model uses district and year fixed effects with standard errors clustered by district. No flypaper effect appears in total expenditure: one additional rupiah of transfers raises spending by 0.866 rupiah against 0.848 rupiah for own-source revenue, a difference of 0.019 with a p value of 0.918. Disaggregation reveals a different picture. Transfers raise capital expenditure by 0.461 rupiah while own-source revenue raises it not at all, with an insignificant coefficient of negative 0.056. The difference of 0.516 carries a p value of 0.010 and survives four alternative specifications. The flypaper effect lies not in the level of spending but in its composition. Transfers fund capital formation while own-source revenue funds recurrent spending.