Budget credibility improved over the period, but with a late reversal. Across FY2019/20–2023/24, TZS 191.846T was approved and TZS 176.372T was spent—91.9% execution—leaving a net underspend of TZS 15.474T. Execution rose from 84.2% (2019/20, lowest) to 97.0% (2022/23, best), then fell by 4.3 percentage points in 2023/24. This pattern is a classic signal of two competing realities: improving in-year delivery capacity through 2022/23, but persistent constraints in a specific financing channel (seen clearly in the local vs forex section) that weakened aggregate execution in the most recent year. For oversight, the important story is not just the execution line—it's which financing components and which votes drive the variance in each year.

This chart is an affordability stress-test rather than a ‘deficit’ calculation. Revenue data is missing for FY2021/22 (shown as N/A, not zero). In years where revenue exists, expenditure averages 98.1% of revenue (range: 91.3% to 104.5%), with FY2022/23 exceeding revenue at 104.5%—a potential signal of financing pressure, timing differences, or reliance on non-revenue financing. For policy makers, persistent years near (or above) 100% mean fiscal space is tight: when execution falls, it may reflect constrained financing and cash management—not only weak implementation. This is why separating local vs forex execution is essential to interpret whether underspending is a ‘capacity problem’ or a ‘financing channel problem.'

Recurrent behaves like a ‘baseline’—development determines whether the budget plan is actually delivered. Over five years, recurrent execution is 96.5% (low volatility: ~3.7pp standard deviation), while development execution is 84.0% (high volatility: ~11.5pp standard deviation). Development also explains most of the national underspend: about TZS 11.247T of the total TZS 15.474T shortfall (~73%). This matters for policy: if the goal is better budget credibility, the biggest gains come from fixing development pipeline readiness (design, procurement, land, safeguards) and the financing channel constraints that disproportionately hit development.

The composition of spending shifts meaningfully—and those shifts change risk. Development’s share of actual spending rises from 28.7% (2019/20) to a peak of 38.3% (2021/22) before settling around ~33–34%. When development’s share rises, overall execution becomes more exposed to delivery and financing risks because development is the volatile component. For decision-makers, this chart helps interpret ‘why execution moved’: a year can show good total execution because recurrent over-performed while development under-delivered, or vice versa—very different policy implications.

This is the single most diagnostic chart on the page. Local-funded execution (recurrent + development local) is strong and stable: 96.2% over five years, reaching ~100% from 2021/22 onward (101.0% in 2022/23 suggests in-year reallocations or accelerated releases). Forex-funded development is the volatility engine: 43.6% execution over five years with very high instability (~21.2pp standard deviation), collapsing to 8.1% in 2023/24. The implication is clear: recent underspending is not primarily a ‘government can’t spend’ story—it's a ‘foreign-financed development isn’t disbursing/being implemented as planned’ story. That points to project readiness, drawdown conditions, procurement, and counterpart/co-financing bottlenecks.

Forex is planned at scale but realized at much smaller scale—creating the national delivery gap. Across five years, approved forex development totals TZS 15.493T but only TZS 6.752T is spent, a forex shortfall of TZS 8.741T. Forex makes up only 22.1% of approved development, yet it accounts for 77.7% of the entire development shortfall—and 56.5% of the total national underspend. FY2023/24 alone contributes ~34.5% of the five-year forex shortfall (a single-year forex gap of ~TZS 3.016T). For policy makers, this is where to look first: strengthening the forex-financed portfolio produces the largest marginal improvement in national execution.

This table measures ‘plan fidelity’ among the biggest budget actors. On average, 9 of the Top10 votes by allocation are also Top10 by spending—so implementation generally follows the approved structure. But divergence increases in the most recent year: in FY2023/24 the overlap drops to 8/10. Two votes were Top10 by allocation but not by spending—Vote 052 (Ministry of Health) and Vote 056 (President Office - Regional Administration and Local Government Authorities)—while two votes moved into the Top10 spenders despite not being Top10 by allocation—Vote 057 (Ministry of Defence and National Service) and Vote 028 (Ministry of Home Affairs-Police Force). For policy oversight, these ‘swaps’ are the key accountability flags: they can reflect deliberate reprioritisation, financing constraints (especially delayed forex/donor development), or implementation bottlenecks in the originally planned high-allocation votes.

Seven of the biggest spenders do not drive execution equally—this heatmap shows which “always-Top10” votes are structurally reliable vs structurally volatile. Only Vote 038 (Defence) stays consistently near plan (~101–115% across the period), while Vote 021 (The Treasury) is persistently under-executed (~28–62% every year), signalling a structural constraint rather than a one-off shock. Volatility is concentrated in a few large votes—Vote 062 (Ministry of Transport) swings from 44% to 128%, Vote 022 (Consolidated Fund Services) from 77% to 138%, and Vote 098 (Ministry of Works) from 89% to 154%—patterns consistent with rephasing/reallocations and financing timing. In FY2023/24, 5 of the 7 votes shown fall below 95% execution, indicating that the latest-year execution drop reflects broad softening across major implementers, not just one isolated ministry.

Interpret these bars as ‘development funding sources’—not all of them are external donors. Source 0GT (Government of Tanzania) dominates development approvals (78.6% of all approved development) and executes strongly over five years (94.4%), meaning Tanzania’s development budget is largely delivered through domestic financing. The policy signal is in the gap between the small number of external/forex-heavy sources and what is actually spent: despite being a minority share of planned development, under-execution by external sources disproportionately drives development underspend and project delays. This pattern points to a planning realism issue: if externally financed components routinely don’t materialize, the approved development envelope overstates what can be delivered unless project readiness and disbursement constraints are fixed.

This trend chart is really ‘domestic vs external reliability’ inside development financing. 0GT (Government of Tanzania) rises from 66.8% execution (2019/20) to 112.2% (2023/24), suggesting domestic development spending has increasingly been used to keep delivery moving—often through reallocations, accelerated releases, or substitution when external financing slows. In contrast, several forex-heavy sources deteriorate sharply by 2023/24 (e.g., 0WB to 1.3%, 0AB to 9.5%). The implication for policy makers is not “donors are volatile” (that’s obvious)—it’s that budget credibility hinges on how conservatively external disbursements are forecast and how quickly stalled externally financed projects can be restructured, retendered, or rephased so the approved development plan matches what is actually financeable and implementable.

This table separates two different execution problems that get mixed in headline ‘development execution’. For 0GT (Government of Tanzania), execution reflects domestic budget management (release decisions, reallocations, absorption capacity). For most other sources—often 100% forex—the execution rate is effectively a measure of disbursement + implementation mechanics (conditions, procurement, safeguards, counterpart funding, and drawdown processing). Among major sources (≥TZS 0.200T approved), execution ranges from 98.3% (0IA) to 17.6% (0GV), and the single largest contributor to negative development variance is 0WB (−TZS 5.000T). For oversight, the ‘Top shortfall votes’ column is the action list: it tells you where external financing is repeatedly not converting into spend. If these shortfalls persist, the policy response is to (i) tighten realism in forex financing projections, and (ii) do portfolio triage—identify which projects are stuck for administrative reasons versus genuinely unready, then rephase or redesign rather than repeatedly approving amounts that won’t execute.