Publication Details
Abstract
Assessments of public-sector reporting quality are normally conducted at the level of a jurisdiction and expressed as the proportion of standards formally adopted. This paper works one level down, at the individual budgetary organisation, and asks what its financial statements are actually capable of telling a reader. We develop three instruments and apply them to the complete 2019-2025 statutory reporting archive of four Uzbek public medical organisations: an index of the structural completeness of the report set, a battery of inter-form articulation tests, and a three-dimensional index of the diagnostic power of analytical indicators. Four results follow. Structural completeness averages 56.3 per cent and ranges from 40.0 to 65.0 across the objects; three criteria score zero everywhere, because the cash-flow statement, the statement of changes in net assets and the notes are not contemplated in the national report set at all. Second, the reliability of a statement depends more on the design of the form than on the competence of the accounting service: the mean absolute relative residual of the internal articulation test fell from 1.798 per cent under the old redaction to 0.0006 per cent under the 2024 redaction, with the same staff, the same registers and the same software. Third, when two forms are populated automatically from a single source their agreement conveys no assurance, and an impossible value of 148,217.2 mn UZS propagated between them unchecked; we call this the automation paradox. Fourth, mean computability of the ten analytical indicators tested is 47.9 per cent, and diagnostic power ranges from 56.0 to 88.1, with the strongest indicators drawn from the balance sheet and the weakest from the expenditure statements. We conclude that arithmetic correctness and economic reliability are distinct properties, and that every material deviation we identified survived seven years of reporting while preserving the accounting equation.