FY 2023/24 spent TZS 41.156T out of TZS 44.388T approved (92.72% execution), leaving a TZS 3.232T underspend. That shortfall is not a ‘small miss’—it is macro-material and signals that financing/implementation constraints dominated the year. Read this chart alongside the Local–Forex section: the headline underspend is overwhelmingly driven by forex-funded development not cash-flowing as planned, not by domestic spending ‘switching off’.

With revenue available for FY 2023/24, actual revenue (TZS 41.808T) slightly exceeded actual expenditure (TZS 41.156T): spending was ~98.44% of revenue. That pattern is consistent with a ‘cash-managed’ execution year—where spending tracks realized inflows rather than the full approved envelope. It should not be interpreted as fiscal surplus by itself (financing/arrears aren’t shown here), but it does suggest that the year’s underspend functioned as an adjustment mechanism when planned funding—especially forex—did not materialize.

Both sides of the budget under-executed in FY 2023/24: recurrent executed at 93.69% (≈TZS 1.841T underspend) and development at 90.86% (≈TZS 1.391T underspend). But the development headline hides a major substitution: development-local over-executed at 113.61% (+TZS 1.624T), while development-forex collapsed to 8.12% (−TZS 3.016T). Policy implication: the key execution risk is not ‘development’ per se—it’s development that depends on volatile forex disbursement pipelines.

Actual spending composition was stable—66.41% recurrent and 33.59% development—very close to the approved split (65.72% vs 34.28%). That stability matters: despite a large financing shock in forex, the government did not dramatically reweight the budget between recurrent and development. Instead, the adjustment happened *inside development* (forex shortfalls partly offset by higher local development execution), which is exactly the kind of shift that can preserve service delivery while quietly rephasing externally financed projects.

Forex was planned to be a meaningful slice of the budget (TZS 3.282T = ~7.39% of total approved), but it delivered only TZS 0.266T—just ~0.65% of total actual spending. Local spending, by contrast, executed at 99.47% with only a small net shortfall (−TZS 0.216T). This is the defining execution story of FY 2023/24: aggregate underspending is primarily a forex funding failure, not a broad domestic implementation slowdown.

Execution is essentially ‘two different worlds’ in FY 2023/24: Local at 99.47% versus Forex at 8.12%. At this magnitude, forex under-execution stops being normal volatility and becomes a pipeline risk signal (procurement delays, disbursement conditions, or project readiness). For oversight, the right question is not only “who underspent?”—it’s “which financing pipelines did not convert approvals into cash and delivery?”

This chart shows the substitution mechanism explicitly. Development-local exceeded plan (113.61%), while development-forex delivered only 8.12% of what was approved. In practical terms, FY 2023/24 looks like a year where domestic resources carried development delivery while many externally financed components stalled. That has implications for future budgets: repeated forex under-execution can create a rolling backlog of ‘approved but undelivered’ projects and distort next-year planning unless the pipeline constraints are fixed.

The biggest votes still dominate the budget: the Top10 by approved account for 64.28% of total allocation, and the Top10 by spending account for 63.34% of total expenditure. Alignment is high but not perfect (8/10 overlap). The two ‘allocation-only’ votes are Vote 052 — Ministry of Health and Vote 056 — President Office - Regional Administration and Local Government Authorities; both under-executed (≈69% and ≈71%) with near-zero forex delivery on sizable approved forex (≈TZS 0.302T and ≈TZS 0.220T shortfalls). The two ‘expenditure-only’ votes are Vote 057 — Ministry of Defence and National Service and Vote 028 — Ministry of Home Affairs-Police Force; Vote 057 in particular shows extreme over-execution (≈971%, +TZS 1.611T), a red-flag pattern that usually reflects major in-year reallocations, reclassification, or supplementary financing. These swaps are where oversight should focus: they show where execution drifted away from the approved structure—and *why*.

This chart is a reality check on development financing in FY 2023/24. Government of Tanzania (0GT) dominates execution: it provided ~78.89% of approved development financing but delivered ~97.41% of actual development spending (≈112% execution), effectively carrying development delivery when external flows under-performed. External sources approved about TZS 3.212T (~21.11% of development approvals) but spent only TZS 0.359T (~2.59% of development actuals)—an execution rate of ~11.17%. The policy implication is blunt: many ‘approved’ development plans were not implementable at scale without reliable external disbursement, forcing the year’s adjustment through project delays rather than through the approved budget design.

The forex story is not only ‘too little was spent’—it’s also ‘who actually delivered changed’. The World Bank pipeline (0WB) accounted for ~45.62% of approved forex (≈TZS 1.497T) but delivered only ~7.39% of actual forex, with ~1.32% forex execution (≈TZS 0.020T). Meanwhile 0IA was planned as a small share (~3.47% of approved forex, ≈TZS 0.114T) but executed at ~99.03% and ended up contributing ~42.27% of actual forex. For policymakers, this is actionable: forex execution risk is concentrated in specific pipelines, and portfolio reliability (not just portfolio size) should shape future project scheduling, contingency planning, and financing diversification.

This table converts donor performance into oversight targets by linking execution outcomes to the votes they finance. FY 2023/24 shows a ‘domestic backstop’ pattern: 0GT over-delivered on development and absorbed delivery pressure, while several large forex-only donors show very low execution—turning approved envelopes into non-delivery. The ‘Top 3 votes financed’ and ‘Top 3 shortfall votes’ columns tell you exactly where each donor’s under-execution lands in the budget (e.g., repeated shortfalls in high-forex votes like Water/Health/Energy). Also note that some donor codes may represent unclassified flows (e.g., “000”), which can produce misleading execution rates when approvals and actuals are recorded under different local/forex classifications—worth flagging as a data-quality and reporting governance issue.