Tax Lab · Macro-fiscal & revenue mobilisation
Budget Credibility & Expenditure Quality Index
A real PEFA 2016 PI-1/PI-2 calculator. Enter approved-vs-actual budget lines and every score below recomputes live, using the exact formulas the prototype's own README specifies.
The formulas are real and faithfully implemented. The pre-filled figures are not Uganda data.
What PEFA PI-1 and PI-2 actually measure
The Public Expenditure and Financial Accountability (PEFA) framework is the standard international yardstick for whether a government's actual spending matches what it approved. PI-1 asks one question at the whole-budget level: did total actual spending deviate from the total approved budget, and by how much. PI-2 asks a sharper question: even if the total matched, did money move between programmes, sectors or ministries away from what was legislated. A government can pass PI-1 while failing PI-2 badly, if it hits its overall spending target only by overspending some lines and underspending others by a lot.
The calculator
Edit the rows below, the approved figures, the actual figures, the labels, or add and remove lines entirely. PI-1, PI-2, the composite Expenditure Quality Index (EQI) and the overall Budget Credibility Score (BCS) all recalculate as you type.
| Budget line | Approved (UGX bn) | Actual (UGX bn) | |
|---|---|---|---|
PI-1: Aggregate expenditure outturn
Approved total 1265bn vs. actual 1207bn.
PI-2: Composition outturn
Weighted composition variance (CV) across budget lines.
EQI: Expenditure Quality Index
Absorption 95 · Composition fidelity 90 · Outcome 50
BCS: Budget Credibility Score
0.35 × PI-1 + 0.30 × PI-2 + 0.35 × EQI, each PI normalised to 0–100.
Formulas taken directly from the prototype's own README (PEFA 2016 PI-1/PI-2, plus its EQI/BCS composite). The rows above are illustrative placeholders, not Uganda figures; replace them with real approved-vs-actual budget lines to get a real score. The outcome-indicator delta is a free input here, not derived from UBOS data, because that linkage isn't implemented in this calculator.
The formulas, exactly
PI-1
deviation_pct = |actual_total − approved_total| / approved_total × 100
Graded A (≤2%) · B (≤5%) · C (≤10%) · D (>10%), scored 4/3/2/1.
PI-2
CV = Σ share_approved_i × (|share_actual_i − share_approved_i| / share_approved_i × 100)
Graded A (≤2.5%) · B (≤7.5%) · C (≤15%) · D (>15%), scored 4/3/2/1.
EQI (0–100)
Equal-weighted average of: Absorption Efficiency (min(actual/approved × 100, 100)), Composition Fidelity (max(100 − PI-2 CV, 0)), and Outcome Improvement (50 + Δ normalised indicator × 5, clamped 0–100).
BCS (0–100)
BCS = 0.35 × norm(PI-1) + 0.30 × norm(PI-2) + 0.35 × EQI
Where norm(PI) = (score − 1) / 3 × 100, converting the 1–4 grade to a 0–100 scale.
Source: PEFA Secretariat (2016), Public Expenditure and Financial Accountability Framework. Formulas and band thresholds taken directly from the Tax Lab prototype's own README (Budget Credibility and Expenditure Quality Index, section 3), designed to run on MoFPED IFMIS/UGATsDB, UBOS, World Bank WDI, EMIS and JMP data; the shipped repo runs on sample budget tables, not a verified official extract.