Title: Commitment fees on undisbursed loans as an indicator of public investment project inefficiency in Uganda
Authors: Dr. Arinaitwe Julius, Akampurira Sarah, Nabaasa Desire
Volume: 10
Issue: 7
Pages: 641-651
Publication Date: 2026/07/28
Abstract:
Public investment in Uganda has been financed substantially through concessional and non-concessional external loans, yet a persistent gap between loan commitments and actual disbursements has meant that government pays commitment fees on capital that remains idle in project accounts. This study examined whether commitment fees charged on undisbursed loan balances could serve as a reliable, quantifiable indicator of inefficiency in the implementation of publicly financed investment projects in Uganda. The study was guided by one main objective and three specific objectives, namely: to describe the pattern of commitment fee exposure across the public investment project portfolio; to establish the relationship between commitment fees and project implementation performance; and to model the extent to which commitment fees, alongside other project-level and institutional factors, predicted overall project efficiency after accounting for clustering of projects within sectors. A cross-sectional analytical design was adopted, drawing on a simulated portfolio of 182 externally financed public investment projects spanning seven sectors, three broad categories of financing source, and four administrative regions of Uganda. Univariate analysis characterised the distribution of loan amounts, disbursement rates, commitment fees, and project efficiency scores; bivariate analysis, comprising one-way analysis of variance and Pearson correlation, examined differences in commitment fee burden across sectors and financing sources and its association with disbursement performance and implementation delay; and a linear mixed-effects model, with a random intercept for sector, was fitted to quantify the contribution of commitment fees to project efficiency net of implementation delay, absorption capacity, and loan size. The analysis found that commitment fees as a proportion of loan value averaged 0.20 percent (SD = 0.28) and were almost perfectly inversely correlated with disbursement rate (r = -0.954, p < 0.001), varied significantly by sector (F = 2.17, p = 0.048) and marginally by financing source (F = 2.95, p = 0.055), and remained a significant negative predictor of project efficiency in the mixed-effects model (? = -7.63, p < 0.001) even after implementation delay, absorption capacity, and loan size were controlled for. The intraclass correlation of 0.365 indicated that more than a third of the unexplained variation in efficiency lay between sectors rather than between individual projects. The study concluded that commitment fees on undisbursed balances were not merely a contractual financing cost but a statistically robust proxy for weaknesses in project readiness, procurement, and absorption capacity that delay disbursement and ultimately erode public investment efficiency in Uganda. It recommended that project preparation and readiness assessment be strengthened before loan effectiveness, that commitment fee exposure be tracked as a routine performance indicator within the Ministry of Finance, Planning and Economic Development's public investment management system, and that disbursement-linked incentives be built into procuring entities' performance contracts to reduce the fiscal cost of undisbursed external financing.