The Unobserved Middle: Absorption and the Identification of Transfer Effects

Abstract

Studies of intergovernmental transfers estimate the effect of money allocated on outcomes achieved. Between those two points sits a stage that is rarely observed: whether the money was spent, and how well. This study argues that the omission is not a data limitation but an identification problem, because an estimate that spans an unobserved stage cannot be attributed to the policy rather than to its implementation. The method is a structured conceptual review of research published between 2017 and 2024 on transfer effects, budget execution and budget credibility. Four propositions follow. The coefficient is a composite of a policy effect and an implementation effect, and the two have different policy implications: one calls for changing the instrument, the other for changing the administration. The composite is not stable across settings, since studies that do observe the administrative stage find it decisive, with one reassignment of tax administration raising collection by about forty-two per cent in the first year and roughly one hundred and nine per cent by year eight without any change in rate or base. Absorption failure is measurable and occurs at scale, with unspent balances documented alongside high fiscal dependency in the same jurisdiction. And budget credibility research shows that the execution stage is driven by political and capacity factors that are unrelated to the transfer being evaluated, which means the omitted variable is not conveniently uncorrelated with the treatment. Absorption should be reported as an intermediate outcome.